\documentclass[twoside]{book} \usepackage{html} % Recent LaTeX distributions include html.sty (written for latex2html). % If yours doesn't, get a copy from CTAN (http://www.ctan.org/), from % the current latex2html package (also available from CTAN), or from % the wave/misc subdirectory of the directory where you found this % file, and put the copy into this directory (or into the directory % that contains the other .sty files referenced below) before attempting % to process this file using LaTeX. \usepackage{epsfig} \usepackage{wrapfig} \usepackage{fancybox} \usepackage{latexsym} \usepackage{makeidx} \usepackage{xspace} \usepackage{verbatim} \makeindex % These commands are based on the fancybox package. Use \keycap{X} to print % a glyph resembling a keyboard `X' key. Use \button{Y} to print a reasonable % facsimile of an Open Look button labelled `Y'. Use \menubutton{Z} to print % a similar Open Look menu button labelled `Z'. Use \amenubutton{X} to print % an abbreviated menu button labelled `X'. \setlength{\fboxsep}{2.25pt} \newcommand{\keycap}[1]{\cornersize{.5}\Ovalbox{\small\sf #1}} \newcommand{\button}[1]{\cornersize{2}\ovalbox{\rule[-.3mm]{0cm}{2.5mm}\small\sf ~#1~}} \newcommand{\menubutton}[1]{\button{#1~\ensuremath{\nabla}}} \newcommand{\amenubutton}[1]{{\sf #1}~\keycap{\ensuremath{\nabla}}} \newcommand{\WAVE}{{\sf WAVE}\xspace} \pagenumbering{roman} \title{\WAVE{} User's Guide} \author{Fifth Edition\\ (revised and with corrections for \WAVE{} version WVRSN)\\ LONGDATE\\ \\ \\ \\ George B. Moody\\ Harvard-MIT Division of Health Sciences and Technology} \date{} \begin{document} \begin{htmlonly} \index{custom menu|see menu file} \index{shell variables|see environment variables} \end{htmlonly} \begin{latexonly} \index{custom menu|see{menu file}} \index{shell variables|see{environment variables}} \end{latexonly} \maketitle \pagestyle{empty} \vspace*{\fill} \noindent Copyright \copyright 1992 -- 2014 George B. Moody \vspace{1 in} \noindent For information on obtaining the most recent version of \WAVE{}, visit \begin{htmlonly} \htmladdnormallink{PhysioNet}{http://physionet.org/}, or write to: \end{htmlonly} \begin{latexonly} PhysioNet ({\tt http://physionet.org/}), or write to: \end{latexonly} \begin{quote} PhysioNet\\ Massachusetts Institute of Technology\\ 77 Massachusetts Avenue, Room E25-505A\\ Cambridge, MA 02139\\ USA\\ \end{quote} \begin{htmlonly} \noindent A \htmladdnormallink{PDF}{wug.pdf} version of this guide is available. The latest version can be downloaded from \htmladdnormallink{MIT}{http://physionet.org/}. \end{htmlonly} \begin{latexonly} \noindent An HTML version of this guide is available at {\tt http://physio\-net.\-org/\-physio\-tools/\-wug/}. \end{latexonly} \vspace{0.2 in} \noindent Permission is granted to make and distribute verbatim copies of this guide provided that the copyright notice and this permission notice are preserved on all copies. % Permission is granted to process this file through TeX and to print the % results, provided that the printed document carries a copying permission % notice identical to this one except for the removal of this paragraph % (this paragraph not being relevant to the printed document). % Permission is granted to process this file through LaTeX2html and to % access the resulting HTML documents and image files using a Web % browser or any other means. \vspace{0.2 in} \noindent Permission is granted to copy and distribute modified versions of this guide under the conditions for verbatim copying, provided also that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. \vspace{0.2 in} \noindent Permission is granted to copy and distribute translations of this guide into another language, under the above conditions for modified versions. \begin{htmlonly} \section*{Other links of interest} \begin{itemize} \item \htmladdnormallink{PhysioNet} {http://physionet.org/} \\ Freely available databases of ECGs and other physiologic signals, and related software. The PhysioNet web site is a public service of the PhysioNet Resource funded by the National Institutes of Health's NIBIB and NIGMS. This is the source for the software described in this guide. \item \htmladdnormallink{WFDB Programmer's Guide}{../wpg/wpg.htm} \\ Includes tutorial and reference material relating to the WFDB library, a portable set of functions (subroutines) for reading and writing files in the formats supported by the applications described here. The WFDB library may be used with C, C++, or Fortran programs; the guide primarily describes the C interface. \item \htmladdnormallink{WFDB Applications Guide}{../wag/wag.htm} \\ Includes {\tt man} pages for the applications included in the WFDB Software Package, specifications for file formats supported by the WFDB library. Detailed installation instructions for the WFDB Software Package are included in an appendix, and a second appendix discusses how the performance of ECG analyzers can be formally evaluated using standard reference databases and the standard evaluation tools included in the WFDB Software Package. \end{itemize} \end{htmlonly} \newpage \setcounter{page}{1} \tableofcontents % Printing the list of figures doesn't work any more -- why?? %\newpage %\listoffigures \newpage \pagestyle{plain} \chapter*{Preface} \addcontentsline{toc}{chapter}{Preface} Many areas of clinical practice and research share a common need for visualization and analysis of physiologic signals. Typically, signals such as electrocardiograms (ECG), respiration, blood pressure, electroencephalograms (EEG), electrooculograms (EOG), and electromyograms (EMG) may be acquired during a clinical procedure or experiment, for durations ranging from a few seconds to many hours. Often these signals must be monitored and analyzed in real time (i.e., while they are being acquired). In other cases, the signals can be recorded for later analysis. There are two common reasons for recording signals for later analysis. First, the analysis may be too complex to perform in real time, or it may require observation of long time periods (for example, in Fourier spectral analysis of very low frequencies). Second, the analytic technique itself, as well as the signals, may be a subject of investigation. Digitally recorded signals are ideally suited as test material in such cases, since they can be used to provide strictly reproducible inputs to a variety of analysis methods. \WAVE{} is a computer program that helps you, the clinician or researcher, to analyze digitally recorded signals. Using \WAVE{}, you can view any desired portion of your signals as if you were browsing through a chart recording. (\WAVE{} can print a paper `chart recording' of any portion of the signals, if you wish.) You can annotate (label) any features of the signals you choose. You can select any subset of the signals, and any time interval, to be analyzed by an external program under the control of \WAVE{}. If the analysis program generates signal annotations, you can view and correct them. Starting with example programs and library functions provided in the WFDB Software Package, you can write your own programs for signal analysis, and add their capabilities to \WAVE{}'s repertoire, without recompiling (or even restarting) \WAVE{}. This guide describes how to use \WAVE{}, how to extend its capabilities, and what hardware and software you will need in order to do so. It does not presume extensive familiarity with computers or signal processing, but it would be helpful to have a friend with some computer experience available when you begin. Many friends have suggested improvements in \WAVE{}. Thanks to Penny Ford-Carleton, Ted Clancy, Kevin Clark, Leon Glass, Scott Greenwald, Farzin Guilak, Jeff Hausdorff, Yuhei Ichimaru, David Israel, Franc Jager, Joe Mietus, Rama Mukkamala, Sheila Ryan, Kambiz Soroushian, Alessandro Taddei, and Andy Wieckiewicz, and to all of the early users of \WAVE{}. I would especially like to thank Roger Mark for his continuous support and encouragement of this project. Your comments on this guide, and on \WAVE{}, are welcome. Please send them to: \begin{quote} PhysioNet\\ MIT Room E25-505A\\ Cambridge, MA 02139\\ USA\\ \\ (e-mail: {\tt wfdb@physionet.org}) \end{quote} \section*{Notes on the third edition} Long time users of \WAVE{} will note that this edition has been greatly expanded over previous editions. In part, this reflects more complete coverage of the original material, but \WAVE{} itself has also grown. \WAVE{} WVRSN is freely available from \hyperref{PhysioNet} {PhysioNet (see section~} {, page~\pageref{sec:getting-wave})} {sec:getting-wave}. I have received many requests for a version of \WAVE{} that runs on a PC. This is possible at last! Under the remarkably complete, and completely free, Linux operating system, \WAVE{} now runs exactly as it does on SPARCstations. If you have wanted to use \WAVE{} but have been limited to using a PC, please try out \hyperref{Linux} {Linux (see section~} {, page~\pageref{sec:linux}, for details)} {sec:linux}. \begin{latexonly} This guide is now available in HTML form as on-line help for \WAVE{} (choose \button{User's Guide} in \WAVE{}'s {\sf Help Topics} window), and it may also be read with any Web browser independently of \WAVE{} (point your browser to {\tt file:\-/usr\-/local\-/help\-/wave\-/wug.htm} if you have installed \WAVE{} on your system, or to\\ {\tt http://physionet.org\-/physiotools\-/wug/}\\ otherwise). \end{latexonly} \begin{htmlonly} This on-line version of the \WAVE{} User's Guide was prepared using \htmladdnormallink{\LaTeX 2\texttt{html}} {http://www.ctan.org/tex-archive/support/latex2html/} from the same \LaTeX{} sources used for the printed manual. You can print your own copy from the PostScript version \htmladdnormallink{here}{wug.ps}. The most recent \LaTeX{}, HTML, PDF, and PostScript versions can be obtained from \htmladdnormallink{PhysioNet} {http://physionet.org/}. I would be grateful for reports of any typographic or other problems in the HTML version, since the translation to HTML is automated and may be less than perfect. Note that the distributed HTML version uses HTML 3.0 table tags, and for this reason the pages containing tables (there are only a few) are best viewed with a browser that can interpret these tags, such as Netscape. If you don't like this feature, an HTML 2.0 version can be produced using \LaTeX 2\texttt{html} (see the file named {.latex2html-init} in the directory of \LaTeX{} sources). \end{htmlonly} Once again, my sincere thanks to all of the users of \WAVE{}, whose questions, comments, and suggestions have helped to make \WAVE{} and this guide more useful for all of us. \vspace{2em} \noindent GBM\\ Cambridge, Massachusetts\\ August, 1996 \section*{Notes on the fourth edition} This interim edition of the \WAVE{} User's Guide is a lightly edited revision of the third edition, changed to reflect the renaming of the WFDB Software Package (formerly known as the DB Software Package) and the related name changes. There have been modest changes in \WAVE{} itself over the past three years (most obviously to the casual user, the main control panel is considerably streamlined to allow more space for the signal window; the illustrations in this edition do not reflect this, however). A substantial revision of this guide, with new illustrations, will be available shortly. \WAVE{} is now free software! It is now part of the WFDB Software Package. The WFDB library sources are now available under the LGPL, and the remainder of the WFDB Software Package, including the sources for \WAVE{}, is under the GPL. As always, I am grateful for reports of any errors or omissions in this guide, and for your comments and suggestions about \WAVE{}. \vspace{2em} \noindent GBM\\ Cambridge, Massachusetts\\ May, 1999 \vfill \section*{Notes on the fifth edition (revised)} This edition contains updated figures, additional tutorial material, and revisions to reflect recent changes in \WAVE{}. The previous release of the \WAVE{} User's Guide announced the GTKWave project, intended to replace the XView-based \WAVE{} with a new implementation based on the GTK+ toolkit. We have redirected our efforts to the original XView-based \WAVE{} since it is now available for all of the popular platforms, while the GTK 1.x project for MS-Windows that inspired development of GTKWave has been abandoned. \index{Open Look} A recent flurry of activity has given new life to the venerable XView toolkit, an open source implementation of the Open Look GUI that \WAVE{} has used since version 1. Notably, XView has been ported to Mac OS X (by Logan Donaldson of York University in Toronto) and to MS-Windows (by Isaac Henry of PhysioNet). As a result, \WAVE{} is now available on these platforms in addition to GNU/Linux, FreeBSD, and Solaris. Meanwhile, an outstanding source of XView documentation has been made available by O'Reilly's Open Books Project (\htmladdnormallink{http://www\-.oreilly\-.com/open\-book/open\-look/} {http://www.oreilly.com/openbook/openlook/}), which provides free on-line copies of the previously unpublished Open Look User's Guide (originally planned as Volume 3 of O'Reilly's X Window System Series), the XView Programming Manual (the out-of-print Volume 7A of the same series), and the XView Reference Manual (Volume 7B, also long out of print). Isaac Henry has also provided hooks for using a mouse's scroll wheel within XView, and future versions of \WAVE{} will take advantage of this feature where available. \vspace{2em} \noindent GBM\\ Cambridge, Massachusetts\\ June, 2002\\ revised June, 2005 \chapter{Introducing \WAVE{}} \pagenumbering{arabic} The rest of this guide will be much easier to follow if you read it while running \WAVE{}, so that you can try out the features described. This chapter contains a five-minute exercise in which you will start \WAVE{}, take a quick look around, and exit from it. The later chapters assume that you have been able to complete this exercise successfully -- if you encounter problems here, get help before continuing. \vspace{5mm} \index{Linux}\index{Solaris}\index{SunOS}\index{X Window System!server} \index{X Window System!client}\index{window manager}\index{Macintosh} \index{Microsoft Windows}\index{SPARCstation}\index{PC} \WAVE{} is an X Window System client application that runs on PCs running GNU/Linux, FreeBSD, or MS-Windows; on Macintoshes running OS X; and on SPARCstations running Solaris or SunOS. X clients communicate with an X server and a window manager, programs that handle display output, and keyboard and mouse input, on behalf of the X clients, in much the same way that the Macintosh and Microsoft Windows operating systems do for applications running on Macs and PCs. Unlike Macintosh and Windows applications, however, X clients need not necessarily run on the same computer as the X server. If your computer can run an X server, and it is connected to a network, you can use it to interact with X clients running on any other computers (``hosts'') connected to the same network. Thus you do not even need to have \WAVE{} running on your own computer in order to use it, since X server software is available for almost all currently manufactured computers (including Macintoshes and Windows PCs) and many older computers as well. \index{WAVE host@\WAVE{} host} You may use X server software running on your own computer to interact with a copy of \WAVE{} running on another networked computer (the \WAVE{} host). (Since the window manager is also an X client, it can run on any computer in the network, although it is usually run on the same computer as the X server.) If your own computer is a Macintosh, PC, or Sparcstation, it can act as the \WAVE{} host, and no network connection is required. If you don't yet have a \WAVE{} host, see \hyperref{System Requirements} {Appendix~} {, page~\pageref{app:system-requirements}} {app:system-requirements} for hardware recommendations. For information about installing \WAVE{}, see \hyperref{Setup} {Appendix~} {, page~\pageref{app:setup}} {app:setup}. If \WAVE{} has already been installed on the \WAVE{} host, proceed to the next section after filling in the worksheet on the next page (get help from an expert such as your system administrator if necessary). \newpage \section{Start-up worksheet for \WAVE{}} \label{worksheet} In order to use \WAVE{}, you must know how to log onto and off of your computer (and the \WAVE{} host, if they are not the same computer). Answers to the questions on this worksheet may also be needed. If you need help, see the notes keyed to each question below, or consult your system administrator. \begin{htmlonly} Print a copy of this page, and record \end{htmlonly} \begin{latexonly} Record \end{latexonly} your answers in the spaces provided for future reference. \vspace{5mm} \index{environment variables!WFDB}\index{WFDB environment variable} \index{environment variables!WFDBCAL}\index{WFDBCAL environment variable} \index{X Window System!server}\index{window manager} \index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} \noindent \begin{tabular*}{\textwidth}{|p{3 in}|r|} \hline { {\bf 1.} What command (if any) must be run \emph{on your computer} to start the X server and the window manager, and open a terminal window? } & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 2.} What command (if any) must be run \emph{on the \WAVE{} host} to initialize the database environment (the {\tt WFDB} and {\tt WFDBCAL} environment variables)?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 3.} If you have a one- or two-button mouse, how do you simulate a middle mouse button click?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 4.} If you have a one-button mouse, how do you simulate a right mouse button click?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline \multicolumn{2}{c}{\emph{Skip the remaining questions if your computer is also the \WAVE{} host.}} \\ \hline { {\bf 5.} What is your computer's name?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 6.} What is the name of the \WAVE{} host?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 7.} What command (if any) must be run \emph{on your computer} to permit the \WAVE{} host to create a window on your computer's display?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 8.} What command must be run \emph{on your computer} to log onto the \WAVE{} host?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline { {\bf 9.} What command (if any) must be run \emph{on the \WAVE{} host} to redirect its output to your computer's display?} & {~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\,} \\ \hline \end{tabular*} \vspace{5mm} \subsubsection*{Notes:} \index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} \begin{description} \item[1. Starting the X server.] On some computers, the X server and the window manager are started automatically whenever you log in, and no command is needed. Otherwise, you must run a command to start the X server before you can run \WAVE{}. This command may be `{\tt startx}' or `{\tt openwin}'; for details, check your X server manual (on a UNIX system, type `{\tt man X}'). If you have a choice of window managers, use {\tt olwm} or {\tt olvwm} if possible \hyperref{(click here for details).} {(see Appendix~} {, page~\pageref{app:system-requirements}).} {app:system-requirements} If no terminal window is available after starting the X server, you can usually open one by moving the mouse pointer over the background (the \emph{root window}) of the display, then clicking the left or right mouse button to open a menu. Depending on your system, the menu may include `{\sf xterm}', `{\sf shell}', `{\sf cmdtool}', `{\sf terminal emulator}', or something similar; choose any of these to open a terminal window. \begin{htmlonly} \index{root window}\index{window!root} \end{htmlonly} \begin{latexonly} \index{root window|emph}\index{window!root|emph} \end{latexonly} \index{terminal window}\index{window!terminal} \item[2. Initializing the environment.] If you are using a recent version of \WAVE{}, you probably will not need to do anything to initialize the environment (the WFDB path and the name of the WFDB calibration file). Unless you keep input files in non-standard places, or if you have created your own calibration file in a non-standard location, you can skip this step. \index{shell}\index{command interpreter} \index{environment variables!SHELL}\index{SHELL environment variable} \index{csh@{\tt csh}}\index{C-shell} \index{cshsetwfdb@{\tt cshsetwfdb} command} \index{setwfdb@{\tt setwfdb} command} \index{profile@{\tt .profile}} \index{login@{\tt .login}} If you do need to initialize your environment, this is typically done using a command at login time. The form of this command depends on what shell (command interpreter) you use on the \WAVE{} host. To identify your shell, log onto the \WAVE{} host \index{WAVE host@\WAVE{} host} and type `{\tt echo \$SHELL}'. If the response contains the characters `{\tt csh}', you are using the C-shell (or a variant of it); in this case, use `{\tt source /usr/bin/cshsetwfdb}' to initialize the environment. Otherwise, use `{\tt .~setwfdb}' to do so (don't omit the `{\tt .}' in this case). Usually, the appropriate command is included in your `{\tt .profile}' or `{\tt .login}' script on the \WAVE{} host, so that it need not be entered each time you log onto the \WAVE{} host. See \htmladdnormallink{{\tt setwfdb(1)}}{../wag/setwfd-1.htm} (type `{\tt man setwfdb}' on the \WAVE{} host) for further information. \item[3. Simulating a middle button click.] On a two-button mouse, this is often done by clicking both buttons simultaneously; on a one-button mouse, this is usually performed by pressing and holding a keyboard key such as \keycap{Shift} while clicking the mouse button. Refer to your X server documentation. \index{mouse!middle button!simulating} \index{middle mouse button!simulating} \index{chording (mouse technique)} \item[4. Simulating a right button click.] This is usually done by pressing and holding a keyboard key while clicking the mouse button. Refer to your X server documentation. \index{mouse!right button!simulating} \index{right mouse button!simulating} \index{chording (mouse technique)} \item[5. Your computer's name.] If you don't know your computer's name, you may be able to discover it by typing the command `{\tt hostname}' (on a UNIX system), or by logging in to the \WAVE{} host and typing the command `{\tt who am I}' (your computer's name should appear at the end of the output). If the \WAVE{} host doesn't recognize your computer by name, use your computer's IP address (in the form {\tt a.b.c.d}, where {\tt a}, {\tt b}, {\tt c}, and {\tt d} are decimal numbers between 0 and 255). \index{hostname@{\tt hostname}} \index{names of computers} \index{WAVE host@\WAVE{} host} \index{who am I command@{\tt who am I} command} \index{IP address} \index{address, IP (Internet Protocol)} \item[6. The \WAVE{} host's name.] If your computer doesn't recognize the \WAVE{} host by name, use the \WAVE{} host's IP address. \item[7. Permitting access to your display.] If you use {\tt ssh} (see the next item) to login to the \WAVE{} host, skip this step. Otherwise, this command is usually needed if your computer is running UNIX. For example, if the name of the \WAVE{} host is {\tt atlantic}, this command would be `{\tt xhost +atlantic}'. If your computer is not running UNIX, there may not be any command required; if in doubt, see your X server manual. \index{xhost command@{\tt xhost} command} \index{ssh command@{\tt ssh} command} \item[8. Logging onto the \WAVE{} host.] This is probably a command of the form `{\tt ssh atlantic}'. You will probably be prompted to enter your password when you execute this command. If you don't have {\tt ssh} on your computer, and an {\tt ssh} server on the \WAVE{} host, it is very strongly recommended that you obtain and install them (both are freely available from \htmladdnormallink{http://www.openssh.org/}{http://www.openssh.org/}). Use {\tt telnet} only as a last resort, and never on a public network. \index{ssh command@{\tt ssh} command} \index{telnet command@{\tt telnet} command} \item[9. Redirecting \WAVE{} output to your display.] \label{sec:setting-display} If you use {\tt ssh} (see above), skip this step. Otherwise, you will need to set the {\tt DISPLAY} environment variable to point to your computer. Assume that the name of your computer is {\tt arctic}. If you use the C-shell on the \WAVE{} host (see note 2 above), the command would be `{\tt setenv DISPLAY arctic:0}'. Otherwise, it would be `{\tt DISPLAY=arctic:0; export DISPLAY}'. It is possible to set up your {\tt .profile} or {\tt .login} script so that this command can be executed automatically each time you log onto the \WAVE{} host. Consult your system administrator for details. \index{environment variables!DISPLAY} \index{SHELL environment variable} \index{profile@{\tt .profile}} \index{login@{\tt .login}} \end{description} \newpage \section{A quick look at \WAVE{}} \index{X Window System!server}\index{window manager}\index{terminal window} To begin this exercise, log onto your computer and do whatever you usually do to start the X server, the window manager, and a terminal window (item 1 from the worksheet). \subsection*{About window managers} { \begin{latexonly} \small \end{latexonly} Some window managers require you to place and size windows manually, so the terminal window may not appear immediately. Usually such window managers show an outline of the window in place of the mouse pointer; typically, you click the left mouse button after dragging the window outline to the desired location on your screen in order to make the window appear. Before going any further, find out how to move windows on your screen if you don't already know how to do so. (You will need to do this occasionally while using \WAVE{}, since its windows may sometimes overlap.) Different window managers have different methods for moving windows; a method that usually works is to press either the left or the center mouse button while pointing to the center section of the window's title bar (at the top edge of the window), and then to move the pointer (which may have been replaced by the window outline) to the desired location before releasing the mouse button. (This common action is called \emph{dragging}, as in the expression `drag the window using the left button'.) Try moving the terminal window now. \index{moving windows}\index{window!moving}\index{dragging} For this exercise, all commands should be typed into the same terminal window. Once again, however, differences in window managers may affect how you do this. Some window managers use a \emph{focus-follows-mouse} policy: you can type into a window whenever the mouse pointer is in the window. Others use a \emph{click-to-type} policy: you must click the left mouse button within a window before you can type into it, but the mouse pointer need not remain in the window while you type (this policy will be familiar to Macintosh and Microsoft Windows users). If you use {\tt olwm} or {\tt olvwm} as your window manager, as recommended, you may choose the policy you prefer using the {\sf Properties} menu (click right on the background or root window, and select {\sf Properties} from the pop-up menu that appears). \index{focus-follows-mouse policy} \index{click-to-type policy} \index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} If you have previous experience with X Window System and Open Look user interface applications, much of what you see will be familiar. This exercise does not assume that you have any such experience, but you may find it helpful to read through an introductory book on the Open Look user interface such as the {\it OpenLook User's Guide}, available freely from (\htmladdnormallink{http://www\-.oreilly\-.com/open\-book/open\-look/} {http://www.oreilly.com/openbook/openlook/}). Please note that, although \WAVE{} uses many Open Look controls, it is not fully Open Look compliant; in particular, actions within its signal window (see chapter 2) do not comply with Open Look style guidelines.} \index{Open Look} \subsection*{If you need to run \WAVE{} remotely} If your computer is also the \WAVE{} host, skip ahead to the next paragraph. Otherwise, type the command(s) from items 7, 8, and 9 of the worksheet, in that order (entering your password if prompted to do so by the \WAVE{} host). These commands are the only differences in the procedures for running \WAVE{} remotely (with different desktop and \WAVE{} host systems) and locally (with your own computer serving as the \WAVE{} host). \index{WAVE host@\WAVE{} host} \subsection*{Checking the environment and the sample record} \label{sec:wfdb-path} In common with all programs built using the WFDB library, \WAVE{} finds its input files in any of the directories specified by the {\em WFDB path} (the value of the {\tt WFDB} environment variable, if set, or a compiled-in default path as given in the WFDB library source file {\tt wfdblib.h} otherwise; see the discussion of the \htmladdnormallink{database path}{../wpg/database-path.htm} in the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm} for details). Verify that the sample record is accessible on the WFDB path using the command ``{\tt wfdbwhich 100s.hea}''. If the response begins with \begin{verbatim} `100s.hea': not found \end{verbatim} \noindent type the command needed to initialize the database environment (item 2 from the worksheet). \index{environment variables!WFDB}\index{WFDB environment variable} In this exercise, you will use \WAVE{} to examine record {\tt 100s}, a sample record containing one minute of annotated two-channel ECG. Record {\tt 100s} is distributed with all versions of \WAVE{} and should be available on any \WAVE{} host. Check that it is available by typing \begin{verbatim} rdsamp -r 100s -f 59.99; rdann -r 100s -a atr -f 59 \end{verbatim} \index{rdsamp command@{\tt rdsamp} command} \index{rdann command@{\tt rdann} command} The output should be \begin{verbatim} 21596 982 995 21597 978 989 21598 975 988 21599 975 989 0:59.508 21423 N 0 0 0 \end{verbatim} If you do not obtain this output, see your system administrator. \subsection*{Starting \WAVE{}} \index{starting WAVE@starting \WAVE{}}\index{wave command@{\tt wave} command} \index{terminal window} \index{WAVE host@\WAVE{} host} \index{environment variables!PATH}\index{PATH environment variable} \WAVE{} is usually started by typing a command in a terminal window. Type: \begin{verbatim} wave \end{verbatim} (Be sure to type this using lower-case letters.) If you see a response similar to `{\tt wave:~Command not found.}', you will need to find the executable copy of \WAVE{} on the \WAVE{} host system. (It may be in a directory that is not part of your {\tt PATH}; if so, you may wish to add that directory to your {\tt PATH} by editing {\tt .cshrc} or {\tt .profile} in your home directory. If the previous sentence makes no sense to you, refer to any introductory book on UNIX for guidance.) If the response is `{\tt wave:~permission denied}', see your system administrator about obtaining permission to run \WAVE{}. If all goes as it should, \WAVE{} prints a concise summary of its command format, which should appear (approximately) as: \begin{verbatim} WAVE version XXX (MMM DD YYYY) WFDB library version XXXXXX (MMM DD YYYY). usage: wave -r RECORD[+RECORD] [ options ] Options are: -a annotator-name Open an annotation file -dpi XX[xYY] Calibrate for XX [by YY] dots/inch -f TIME Open the record beginning at TIME -g Use shades of grey only -H Use high-resolution mode -m Use black and white only -O Use overlay graphics -p PATH Search for input files in PATH (if not found in the WFDB path) -s SIGNAL [SIGNAL ...] Initialize the signal list -S Use a shared colormap -Vx Set initial display option x wave is an X11 client. You must specify the X server connection for it in the DISPLAY environment variable. Be sure to set the WFDB environment variable to indicate a list of directories that contain input files for wave. For more information, type `more /usr/help/wave/wave.hlp', or open `http://physionet.org/physiotools/wug/' using your web browser. \end{verbatim} \noindent The comments about setting the {\tt DISPLAY} and {\tt WFDB} variables will not appear if you have already set these variables. \index{environment variables!WFDB}\index{WFDB environment variable} \index{environment variables!DISPLAY}\index{DISPLAY environment variable} Now try using \WAVE{} to view record {\tt 100s}, together with its {\tt atr} (reference) annotations. Type: \begin{verbatim} wave -r 100s -a atr & \end{verbatim} The `{\tt \&}' is optional but recommended; it allows you to type additional commands in the terminal emulator window without interrupting or exiting from \WAVE{}. After a few seconds, \WAVE{}'s \emph{main window} (figure~\ref{fig:main-window}) opens. \index{atr (reference) annotations@{\tt atr} (reference) annotations} \index{annotation!reference}\index{reference annotations} \begin{htmlonly} \index{main window} \end{htmlonly} \begin{latexonly} \index{main window|emph} \end{latexonly} \begin{figure} \centerline{\epsfig{file=main-window}} \caption{\WAVE{}'s main window.} \begin{htmlonly} \index{main window} \end{htmlonly} \begin{latexonly} \index{main window|textbf} \end{latexonly} \label{fig:main-window} \end{figure} \subsection*{A Five-Minute Tour of \WAVE{}} \begin{htmlonly} \index{title bar}\index{window!title bar} \index{main control panel} \end{htmlonly} \begin{latexonly} \index{title bar|emph}\index{window!title bar|emph} \index{main control panel|emph} \end{latexonly} At the top of the window is the \emph{title bar}, which reads `{\sf Record 100s atr}'. Below the title bar is \WAVE{}'s \emph{main control panel}, which contains three groups of buttons. Several of these (\menubutton{File}, \menubutton{Edit}, and \menubutton{Properties}) are \emph{menu buttons} (distinguished by the `$\nabla$' at the right end of the button), and are selectable using the right mouse button. The other buttons are selected using the left mouse button. \begin{htmlonly} \index{signal!window} \index{time indicator} \end{htmlonly} \begin{latexonly} \index{signal!window|emph} \index{time indicator|emph} \end{latexonly} \index{rhythm label}\index{annotation!rhythm} Below the main control panel is the \emph{signal window}. This window shows a portion of the sample record {\tt 100s}, which contains two signals. Between the two signals, \emph{annotations} are shown. The first annotation (`{\sf (N}', at the left edge of the window) is shown below the level of the others, to indicate that it is a rhythm label (`{\sf (N}' means normal sinus rhythm). Most of the other annotations are `{\sf N}', and indicate normal beats; the `{\sf A}' indicates an atrial premature beat. In the lower corners of the signal window, the \emph{time indicators} (`{\sf 0:00}' and `{\sf 0:10}') show the elapsed time in minutes and seconds from the beginning of the record to the samples at the left and right edges of the window respectively. (If your display is less than about 260 mm wide, the signal window will not show a full 10 seconds of the record.) When you move the pointer into the signal window, it changes shape. Annotation editing and other operations are possible while the pointer is in the signal window; these operations are described in later exercises. \index{controls!navigation}\index{navigation controls} \index{moving through a record} The buttons in the middle group (\button{{\tt <} Search}, \button{\tt <<}, \button{\tt <}, \button{Find...}, \button{\tt >}, \button{\tt >>}, and \button{Search {\tt >}}) are the controls that you use to navigate through the record. The \button{\tt >>} button advances the signal window display by the width of the window (10 seconds in this case). Move the mouse pointer over \button{\tt >>} and press the left button. (This common action is referred to below as `clicking left on \button{\tt >>}'.) The signal window is redrawn, and now shows the interval from 0:10 to 0:20. The \button{\tt >} button also advances the window, but by only half the window width. Try it now. Similarly, \button{\tt <<} moves the window back by one screenful, and \button{\tt <} moves it back by half a screenful. Use these buttons to move to a window that begins at 0:55. Since the record is only 1 minute long, the right half of the window is empty. Notice that you can continue to advance past the end of the record if you wish. (\WAVE{} will not allow you to back up past the beginning of the record, however.) \index{pop-up windows} Buttons such as \button{Find...}, with labels that end with `{\sf ...}', make \WAVE{} open other (so-called `pop-up') windows. Click left on \button{Find...} now. The {\sf Find} window (figure~\ref{fig:find-window}) will appear. \begin{figure} \centerline{\epsfig{file=find-window}} \caption{The {\sf Find} window.} \begin{htmlonly} \index{Find window@{\sf Find} window} \end{htmlonly} \begin{latexonly} \index{Find window@{\sf Find} window|textbf} \end{latexonly} \label{fig:find-window} \end{figure} Its location on your screen is controlled by your window manager; typically it will appear either at the upper-left corner of the screen, or directly beneath the pointer (obscuring part of the signal window). If you wish, move the {\sf Find} window to another location on your screen. \begin{htmlonly} \index{text cursor} \index{insertion point} \index{keyboard focus} \end{htmlonly} \begin{latexonly} \index{text cursor|emph} \index{insertion point|emph} \index{keyboard focus|emph} \end{latexonly} Within the {\sf Find} window are four text fields in which you may type. Move the pointer into the the part of the window below the title bar. A black, upward-pointing triangle (the text cursor or \emph{insertion point}) should appear in one of the three fields. (Once again, your window manager may influence what you see. If you see a grey diamond rather than a black triangle, you must perform whatever action your window manager requires to give the \emph{keyboard focus} to the {\sf Find} window. Usually, clicking left is sufficient.) \index{start time@{\sf Start time}} \index{end time@{\sf End time}} The {\sf Start time} and {\sf End time} fields match the times shown in the lower corners of the signal window. If you change either of these fields, the signal window moves accordingly. To change a text field, move the pointer to the right of the existing text and click left, press \keycap{Backspace} or \keycap{DEL} to erase any characters you wish to change, and type in the desired value, finishing by pressing \keycap{Enter} (or \keycap{Return}). (Always remember to press \keycap{Enter} after changing a text field; your changes are not registered until you press \keycap{Enter}.) Try changing the {\sf Start time} and {\sf End time} fields now, and watch how the contents of the signal window change. You can type any desired time into these fields; specifically, it need not be a multiple of 5 seconds. \index{search}\index{annotation!searching for} Using any of the controls you have seen so far, set the signal window so that it shows the segment of the record beginning at 0:22. Now select the {\sf Search for annotation} field, and enter `{\tt A}' (followed, as always, by \keycap{Enter}). This action asks \WAVE{} to find the next occurrence of an `{\sf A}' (atrial premature beat) label, and to redraw the signal window roughly centered on that label. As you may have noticed, the only `{\sf A}' annotation in record {\tt 100s} occurs at about 0:06, so \WAVE{} is unable to satisfy this request. Whenever \WAVE{} cannot do what you have asked, it displays \begin{wrapfigure}[8]{r}{3cm} \mbox{\epsfig{file=nomatch}} \begin{htmlonly} \index{notice box} \end{htmlonly} \begin{latexonly} \index{notice box|textbf} \end{latexonly} \end{wrapfigure} a \emph{notice box}, with a brief message describing what has happened. (In this case, the message is `{\sf No match found!}', in other words, there are no `{\sf A}' labels between 0:32 and the end of the record.) Notice boxes are usually accompanied by a beep (although your window manager may allow you to disable the beep). Notice boxes always contain at least one button (in this case, labelled \button{Continue}). When a notice box is displayed, you \emph{must} click left on one of the buttons in the notice box before you can do anything else in \WAVE{}. Click left on \button{Continue} now to dismiss the notice box. \index{search}\index{annotation!searching for} The \button{{\tt <} Search} and \button{Search {\tt >}} buttons on \WAVE{}'s main control panel are used to search for annotations that match the {\sf Search for} field in the Find window. Searching forward using \button{Search {\tt >}} will be unsuccessful (we know this, because \WAVE{} has already tried to do so after we changed the contents of the {\sf Search for} field). Use \button{{\tt <} Search} button to search backwards. \begin{wrapfigure}{l}{2.2cm} \mbox{\epsfig{file=file-print}} \end{wrapfigure} The signal window now shows the segment from 0:01 to 0:11, with the `{\sf A}' annotation at 0:06 roughly centered in the window. If you use either \button{{\tt <} Search} or \button{Search {\tt >}} now, the search fails, because the only `{\sf A}' annotation in the record is already on-screen. For more information, see \hyperref{Searching for annotations} {section~} {, page~\pageref{sec:searching}} {sec:searching}. \index{printing!signal window contents}\index{screen dump} \index{PostScript} If a PostScript printer (or a PostScript interpreter such as GhostScript) is available, try printing the contents of the signal window. To do this, press and hold the right mouse button while the pointer is above \menubutton{File}; then drag the pointer downwards until the {\sf Print} selection is highlighted, and release the right mouse button. Most modern printers will be able to print the page at (nearly) full speed; older PostScript printers may require several minutes to do so. It is not necessary to wait for the output to appear before continuing. The output will appear as in figure~\ref{fig:chart1}. \begin{figure} \fbox{\centerline{\epsfig{file=chart1,width=.9\linewidth}}} %\fbox{\centerline{\epsfig{file=chart1,angle=-90,width=.9\linewidth}}} % This figure was produced in ``landscape'' mode. To print it % properly in portrait mode, it must be rotated by -90 degrees % *before* scaling it. \caption{Printed ``chart recording'' made using {\sf Print} from the {\sf File} menu.} \label{fig:chart1} \end{figure} \begin{htmlonly} \index{spot help}\index{Open Look!spot help} \index{help window} \end{htmlonly} \begin{latexonly} \index{spot help|emph}\index{Open Look!spot help|emph} \index{help window|emph} \end{latexonly} \index{help}\index{on-line help} \index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} \WAVE{} has three types of on-line help. The first type, called \emph{spot help}, works properly only if you are using an Open Look window manager ({\tt olwm} or {\tt olvwm}), although a local expert may be able to show you how to use it with another window manager. To use spot help with an Open Look window manager, move the pointer to any control or display area in a \WAVE{} window and press the \keycap{HELP} key (or the \keycap{F1} key if your keyboard does not have a \keycap{HELP} key). A \emph{spot help window} (see figure~\ref{fig:spot-help}) appears; it contains a magnifying glass icon showing the area you have selected, and a description of the control or display. \begin{figure} \index{spot help} \begin{htmlonly} \index{Open Look!spot help} \end{htmlonly} \begin{latexonly} \index{Open Look!spot help|textbf} \end{latexonly} \label{fig:spot-help} \centerline{\epsfig{file=spot-help}} \caption{A spot help window.} \end{figure} \index{printing!on-line help} \index{help!printing} \index{window!menu button}\index{closing windows} To get an overview of how to use \WAVE{}, or if you cannot use spot help, click left on \button{Help} in \WAVE{}'s main control panel. \begin{figure} \centerline{\epsfig{file=help-topics}} \caption{The {\sf Help Topics} window.} \begin{htmlonly} \index{Help Topics window@{\sf Help Topics} window} \end{htmlonly} \begin{latexonly} \index{Help Topics window@{\sf Help Topics} window|textbf} \end{latexonly} \label{fig:help-topics} \end{figure} The {\sf Help Topics} window (see figure~\ref{fig:help-topics}) appears, from which you may select any of the available topics by clicking left on the corresponding button. If you do so, a scrollable \emph{help text window} (see figure~\ref{fig:help-text}) appears, containing information on the selected topic. \begin{figure} \centerline{\epsfig{file=help-intro}} \caption{A help text window.} \begin{htmlonly} \index{help text window}\index{window!help text} \end{htmlonly} \begin{latexonly} \index{help text window|textbf}\index{window!help text|textbf} \end{latexonly} \label{fig:help-text} \end{figure} Scroll up and down through the window by clicking left on the black triangles in the scroll bar (or use any other Open Look methods if you are familiar with them). If you wish, you can click left on the \button{Print} button at the top of the window to obtain a paper copy. Dismiss the window by clicking right on the \emph{window menu button} (the square button containing a triangle at the upper left corner of the window), and then clicking left on the {\sf Quit} item in the window menu. \index{Mozilla}\index{web browsers} \index{on-line manual}\index{User's Guide (on-line)} \index{WAVE User's Guide (on-line)@\WAVE{} User's Guide (on-line)} The third (and most comprehensive) form of on-line help can be used if a suitable web browser is available on the \WAVE{} host system. (\WAVE{} uses Mozilla by default.) Set the environment variable {\tt URLV} to the name of your browser if you prefer a different browser. See \hyperref{\WAVE{} and the Web} {section~} {, page~\pageref{sec:web},} {sec:web} for details on configuring \WAVE{} and your browser to work together. Click left on the \button{WAVE User's Guide} button in the {\sf Help Topics} window to open this manual using your web browser. If your browser is not running already, this may take a few moments while \WAVE{} starts it up. \index{Quit button@{\sf Quit} button} \index{WAVE host@\WAVE{} host} \begin{wrapfigure}[7]{r}{4.75cm} \mbox{\epsfig{file=notice-quit}} \end{wrapfigure} To complete this exercise, exit from \WAVE{} by clicking left on \button{Quit} in \WAVE{}'s main control panel. A notice box appears as shown at right; click left on \button{Confirm} to exit, or on \button{Cancel} if you're having so much fun that you don't want to stop yet. \chapter{Annotation Editing} There are five editing operations that you can perform using \WAVE{}: changing, moving, copying, inserting, and deleting annotations. In this chapter, you will create and edit an annotation file using \WAVE{} and an external QRS detection program. Start as in the previous chapter, but use the command \begin{verbatim} wave -r 100s & \end{verbatim} \noindent omitting the `{\tt -a atr}' option this time. Note that the title bar contains only the record name. \begin{wrapfigure}[9]{l}{2.2cm} \mbox{\epsfig{file=file-analyze}} \end{wrapfigure} Click right on \menubutton{File}, and click left on the {\sf Analyze...} selection in the File menu. After a short delay, the {\sf Analyze} window (see figure~\ref{fig:analyze-window}) and the {\sf Analysis Commands} window (see figure~\ref{fig:analysis-commands}) appear. \begin{figure} \centerline{\epsfig{file=analyze-window}} \caption{The {\sf Analyze} window.} \begin{htmlonly} \index{Analyze window@{\sf Analyze} window} \end{htmlonly} \begin{latexonly} \index{Analyze window@{\sf Analyze} window|textbf} \end{latexonly} \label{fig:analyze-window} \end{figure} At any time, you can click on \button{Show command window} in the {\sf Analyze} window to make the {\sf Analysis Commands} window visible if it becomes covered by another window. Below the title bar of the {\sf Analyze} window, text fields allow you to specify the beginning and end of a segment of the record, and to choose signals. The {\sf Analyze} window also contains buttons specifying various actions that you can apply to the signals. The separate {\sf Analysis Commands} window behaves like an OpenWindows {\tt cmdtool} window (a scrollable terminal emulator). \index{cmdtool@{\tt cmdtool}} \begin{figure} \centerline{\epsfig{file=analysis-commands}} \caption{The {\sf Analysis Commands} window.} \begin{htmlonly} \index{analysis commands window@{\sf Analysis Commands} window} \end{htmlonly} \begin{latexonly} \index{analysis commands window@{\sf Analysis Commands} window|textbf} \end{latexonly} \label{fig:analysis-commands} \end{figure} Commands may be run in the {\sf Analysis Commands} window either by clicking left on action buttons \index{action buttons} in the {\sf Analyze} window, or by typing directly into the window (depending on your window manager's policy for moving the keyboard focus, you may need to click left within the window before you can type into it, however). The {\sf Analysis Commands} window may be resized if desired, and standard Open Look user interface methods may be used to cut and paste text between it and other windows, including those in other applications. \index{annotation file!creating!using an external program} \index{creating an annotation file!using an external program} To create an annotation file, click left on \button{Mark QRS complexes}. The command \begin{verbatim} sqrs -r 100s -f 0 -t 1:00.000 -s 0 \end{verbatim} \noindent appears in the {\sf Analysis Commands} window, as shown in figure~\ref{fig:analysis-commands}. Below the command, its error output appears. When the command has run to completion (which will take only a few seconds unless the \WAVE{} host is heavily loaded), an annotation file named `{\tt 100s.qrs}' is written into the current directory. If you wish, the {\sf Analyze} window or the {\sf Analysis Commands} window, or both, may be dismissed (by selecting {\sf Quit} from the window menu to the left of the title bar); this can be done at any time, even while analysis is still in progress. If you recall the {\sf Analysis Commands} window later (using \button{Show command window} in the {\sf Analyze} window), it will show any output sent to it while it was dismissed. Note that the {\sf Analysis Commands} window is dismissed whenever you exit from \WAVE{}, and its contents are not retained after that point. \section{Loading annotations to be edited} \index{title bar} \index{annotator name} \index{calibration file} \index{database path} \begin{wrapfigure}[8]{l}{2.2cm} \mbox{\epsfig{file=file-load}} \end{wrapfigure} To load annotations \index{annotation file!loading}\index{loading an annotation file} from `{\tt 100s.qrs}' into \WAVE{}, click right on \menubutton{File}, and click left on the {\sf Load...} selection in the File menu. \begin{figure} \centerline{\epsfig{file=load-window}} \caption{The {\sf Load} window.} \begin{htmlonly} \index{Load window@{\sf Load} window} \end{htmlonly} \begin{latexonly} \index{Load window@{\sf Load} window|textbf} \end{latexonly} \label{fig:load-window} \end{figure} The {\sf Load} window (see figure~\ref{fig:load-window}) appears. It contains text items that specify a record and and an annotator name. Click left next to the {\sf Annotator} field, type `{\tt qrs}', and, as always, press \keycap{ Enter}. The contents of `{\tt 100s.qrs}' are loaded into \WAVE{}, and annotations appear in the signal window. The title bar changes to indicate the annotator name. The {\sf Load} window also contains items that specify the name of the calibration file and the \hyperref{database path} {database path (see section~} {, page~\pageref{sec:wfdb-path});} {sec:wfdb-path} you will not need to change these items for this exercise. You may dismiss the {\sf Load} window or leave it up as you wish. If you have looked through the record using the navigation buttons, use them now to return to the beginning of the record. The time indicator in the lower left corner of the signal window should read `{\sf 0:00}'. Since the \button{Mark QRS complexes} button only produces `{\sf N}' (normal) beat labels, the eighth beat in the window (an atrial premature beat) is mislabelled. In the next part of this exercise, we will correct this label. \section{The {\sf Annotation Template}} Move the pointer into the signal window and click left. \WAVE{}'s {\sf Annotation Template} window appears (see figure~\ref{fig:annotation-template}). \begin{figure} \centerline{\epsfig{file=annotation-template}} \caption{The {\sf Annotation Template} window.} \begin{htmlonly} \index{Annotation Template window@{\sf Annotation Template} window} \end{htmlonly} \begin{latexonly} \index{Annotation Template window@{\sf Annotation Template} window|textbf} \end{latexonly} \label{fig:annotation-template} \end{figure} The contents of the {\sf Annotation Template} window indicate what type of annotation will be inserted if you decide to insert an annotation. Click right on the \amenubutton{Type:} abbreviated menu button. The {\sf Type} menu \index{Type menu@{\sf Type} menu} appears (see figure~\ref{fig:type-menu}). \begin{figure} \centerline{\epsfig{file=type-menu}} \caption{The {\sf Type} menu.} \label{fig:type-menu} \begin{htmlonly} \index{Type menu@{\sf Type} menu} \end{htmlonly} \begin{latexonly} \index{Type menu@{\sf Type} menu|textbf} \end{latexonly} \end{figure} Select `{\sf A~Atrial premature beat}' from the {\sf Type} menu by clicking left on it. This action dismisses the {\sf Type} menu, and your selection now appears in the {\sf Annotation Template} window to the right of the \amenubutton{Type:} menu button. The other fields of the {\sf Annotation Template} can be ignored for this exercise. Single-character annotation mnemonics \index{annotation!mnemonic} that appear in the {\sf Annotation Template}'s {\sf Type} menu can be used as shortcuts \index{editing shortcuts} during annotation editing. (These mnemonics are the same as those used by \WAVE{} to indicate annotations in the signal window.) Instead of calling up the {\sf Type} menu, you may simply type the mnemonic at any time while the pointer is in the signal window, provided that the {\sf Annotation Template} window is open. If you type a recognized mnemonic, the {\sf Annotation Template} is updated immediately to reflect your choice. \index{annotation!enabling editing} \begin{latexonly} \newpage \end{latexonly} \begin{wrapfigure}[8]{r}{6.3cm} \mbox{\epsfig{file=noedit}} \end{wrapfigure} Annotation editing is disabled by default, to avoid inadvertent modification of annotation files. If you now move the pointer into the signal window and click the middle button, this action is understood to mean that you wish to insert an annotation of the type shown in the {\sf Annotation Template} at the location indicated by the pointer. Since \WAVE{} cannot satisfy this request until annotation editing has been enabled, a notice box (shown above) appears to warn you that no action has been taken. Click left on \button{Continue} to dismiss the notice box. \begin{wrapfigure}{l}{3cm} \mbox{\epsfig{file=allow-edit}} \begin{htmlonly} \index{Edit menu@{\sf Edit} menu} \end{htmlonly} \begin{latexonly} \index{Edit menu@{\sf Edit} menu|textbf} \end{latexonly} \end{wrapfigure} To enable annotation editing, click right on \menubutton{Edit}, and click left on the {\sf Allow editing} selection in the {\sf Edit} menu. (Your selection from the {\sf Edit} menu persists for the duration of your {\sf WAVE} session, unless you return to the {\sf Edit} menu to change it. Thus it is sufficient to enable annotation editing once if you plan to edit several records in a single session.) \section{Selecting an annotation} \label{sec:selecting-annotation} \begin{htmlonly} \index{selection rectangle} \index{marker bars} \end{htmlonly} \begin{latexonly} \index{selection rectangle|emph} \index{marker bars|emph} \end{latexonly} \index{annotation!selecting} In this case, we need to change an existing `{\sf N}' annotation into an `{\sf A}' annotation. To do so, we must select the annotation to be changed. Annotations are selected by clicking left or right in the signal window. Whenever the pointer is in the signal window, clicking left selects the annotation to the left of the pointer, and clicking right selects the annotation to the right of the pointer. (If `Num Lock' is off, and the pointer is within the signal window, you can use the \keycap{$\leftarrow$} or \keycap{$\rightarrow$} keys instead of the mouse buttons if you prefer.) The annotation is highlighted by a \emph{selection rectangle} drawn around it, and the pointer jumps to the center of the rectangle (see figure~\ref{fig:main-with-markers}). \begin{figure} \centerline{\epsfig{file=main-with-markers}} \caption{The main window, showing the selection rectangle and marker bars.} \begin{htmlonly} \index{selection rectangle} \index{marker bars} \end{htmlonly} \begin{latexonly} \index{selection rectangle|textbf} \index{marker bars|textbf} \end{latexonly} \label{fig:main-with-markers} \end{figure} If clicking left or right would otherwise move the pointer out of the signal window, \WAVE{} recenters the signal window on the selected annotation. While any of the mouse buttons is down, \emph{marker bars} appear above and below the pointer to allow you to identify the pointer location with respect to the signals. It is often useful, especially while editing annotations, to display marker bars above and below \emph{each} annotation. If you wish to do so, click left on \button{View...}, then click left on the \keycap{markers} selection in the {\sf View} window that appears (see figure~\ref{fig:view-window}), and finally click left on \button{Redraw} to dismiss the {\sf View} window and redraw the signal window with annotation marker bars (see figure~\ref{fig:main-with-markers}). \begin{figure} \centerline{\epsfig{file=view-window}} \caption{The {\sf View} window.} \label{fig:view-window} \begin{htmlonly} \index{View window@{\sf View} window} \end{htmlonly} \begin{latexonly} \index{View window@{\sf View} window|textbf} \end{latexonly} \end{figure} The marker bars can be turned off by repeating this procedure. \section{Changing an annotation without moving it} \index{annotation!changing without moving} \index{changing an annotation} \index{annotator name} Select the eighth beat label by clicking left or right, or using the arrow keys, until the selection rectangle surrounds it. Change the `{\sf N}' to an `{\sf A}' by clicking the middle button or pressing \keycap{F2}. (On some keyboards, if `Num Lock' is off, you can use the \keycap{5} key on the numeric keypad instead of \keycap{F2} if you prefer.) The `{\sf N}' is \begin{wrapfigure}[4]{r}{3cm} \mbox{\epsfig{file=title-with-parens}} \begin{htmlonly} \index{title bar!parentheses in} \end{htmlonly} \begin{latexonly} \index{title bar!parentheses in|textbf} \end{latexonly} \end{wrapfigure} replaced by an `{\sf A}' and the selection rectangle disappears. To indicate that there are unsaved edits in the annotation file, the annotator name in \WAVE{}'s main window title bar is marked by parentheses. \section{Moving an annotation} \index{annotation!moving} \index{moving an annotation} Although no further editing should be necessary, try out the other editing operations now. To move an annotation, select it with the left or right button, drag its marker bars (with the left or right button still depressed) to the desired location, then release the mouse button to drop the annotation. The selection rectangle stays in the original location until the button is released, but the marker bars move with the pointer once the pointer moves outside of the rectangle. To move an annotation by less than the width of the rectangle, simply drag the pointer above or below the rectangle. \index{keyboard commands} If you prefer, you may use the keyboard rather than the mouse for this operation. To do so, drag the marker bars left or right using \keycap{F3} or \keycap{F4}, and then drop the annotation using \keycap{F2}. If you change your mind about moving an annotation, drag the pointer back into the selection rectangle, and then drop the annotation. \section{Inserting an annotation} \index{annotation!inserting}\index{inserting an annotation} To insert an annotation, move the pointer into the signal window (but outside of the selection rectangle if it is visible), press the middle button, drag the marker bars to the desired location, and release the middle button to complete the insertion. \section{Copying an annotation} \index{annotation!copying}\index{copying an annotation} To copy an annotation, select it, then press the \keycap{Copy} key (or \keycap{F6} if your keyboard does not have a \keycap{Copy} key). This action copies the selected annotation into the {\sf Annotation Template}. Now insert the copy at the desired location. (This is a particularly useful shortcut if you are working with complex annotations for which fields other than the {\sf Type} must be set.) \section{Deleting and restoring annotations} \index{annotation!deleting}\index{deleting an annotation} \index{annotation!restoring deleted}\index{restoring a deleted annotation} \index{undeleting an annotation}\index{finding deleted annotations} \index{locating deleted annotations}\index{search!for deleted annotations} To delete an annotation, first click right on the {\sf Annotation Template} window's \amenubutton{Type:} menu button, and select `{\sf . (deleted annotation)}', the first entry. Now select the annotation to be deleted by clicking left or right in the signal window, and click the middle button to delete it. The annotation is replaced by a `{\sf .}' marker, which remains in \WAVE{}'s annotation buffer until a new annotation file is read or until you exit from \WAVE{}. This feature allows you to locate deleted annotations if necessary (you may search for them by entering `{\tt .}' in the {\sf Search for} field of the {\sf Find} window; see \hyperref{Searching for annotations} {section~} {, page~\pageref{sec:searching}} {sec:searching}). To restore a deleted annotation, simply repeat the operations that you used to delete it. \section{Markers} The `{\sf .}' that appears when you delete an annotation is an example of a \emph{marker}. \begin{htmlonly} \index{marker} \end{htmlonly} \begin{latexonly} \index{marker|emph} \end{latexonly} The important difference between \emph{markers} and \emph{annotations} \index{annotation!vs. marker} is that markers are never written to annotation files; they remain in \WAVE{}'s annotation buffer only until \WAVE{} reads another annotation file, or until you exit from \WAVE{}, whichever comes first. Other types of markers are `{\sf :}' (index marks, \index{index mark} which you may insert anywhere in a record, and which you can return to by searching for them), `{\tt <}' (beginning of region), and `{\tt >}' (end of region). \index{beginning of region}\index{end of region}\index{region of interest} The `{\tt <}' and `{\tt >}' markers are special in that only one of each may be defined at any time; inserting a new one deletes the old one, if any. If the {\sf Analyze} window \index{Analyze window@{\sf Analyze} window} is visible, you will notice that inserting or moving the `{\tt <}' marker has the effect of changing the contents of the {\sf Start (elapsed)} field (and the {\sf From} field, if it is enabled); another way to insert a `{\tt <}' marker is to enter the desired time directly into the {\sf Start (elapsed)} field. The `{\tt >}' marker is similarly coupled with the {\sf End (elapsed)} field (and the {\sf To} field) of the {\sf Analyze} window. \section{Changing the {\sf Annotation Template}} \WAVE{} maintains a record of several of the most recently used settings in the {\sf Annotation Template} (this information is not saved when you exit \WAVE{}, however). You can cycle through these settings if you have a three-button mouse; to do so, press and hold the middle button as if to insert an annotation, then click left or right to change the contents of the {\sf Annotation Template}. (If you don't wish to insert an annotation, release the middle button before releasing the left or right button.) This feature is particularly useful if you are manually annotating a record and you find it necessary to use a small number of different annotations to do so. \section{Changing many annotations at once} \index{annotation!changing many at once}\index{changing many annotations} \index{shortcut} The usual method for changing annotations (clicking right or left to select the annotation to be changed, then clicking the middle button to perform the change) can become tedious if you have many annotations to change. \WAVE{} offers a shortcut if you wish to change all annotations in a specified region in the same way. First, specify the region by inserting a `{\tt <}' marker before the first annotation to be changed, and a `{\tt >}' marker after the last annotation to be changed. Next, fill in the {\sf Annotation Template} to indicate how you wish to change the annotations in the region. Finally, click left on \button{Change all in range}. (If the region includes any annotations that are not visible on-screen, \WAVE{} prompts you to confirm that you wish to change them.) Try out this technique now. Insert the `{\tt <}' marker at 0:15, the `{\tt >}' marker at 0:30, and change all of the annotations in the region to type `{\sf V}' (premature ventricular contraction). Inspect the results, then change them back to `{\sf N}' annotations. \section{Searching for annotations} \label{sec:searching} \WAVE{} can search in either direction for the next annotation that meets criteria you specify using the {\sf Search for} field in the {\sf Find} window. Any string that \WAVE{} displays as an annotation label can be entered into the {\sf Search for} field as a search criterion; these strings include annotation and marker mnemonics (usually single characters; see the {\sf Type} menu for a list), signal quality strings (these contain one or more characters from the set ``{\tt c}'', ``{\tt n}'', and ``{\tt u}''; ``{\tt U}'' is also a legal signal quality annotation), and {\em any substring} within a note, rhythm change, ST or T change, or link annotation. In addition to these, there are several {\sf Search for} strings that \WAVE{} interprets in special ways: \begin{tabular*}{5.5 in}{l p{4.75 in}} {\tt *v} & {matches any ventricular ectopic beat} \\ {\tt *s} & {matches any supraventricular ectopic beat} \\ {\tt *n} & {matches any other beat type} \\ {\tt *} & {matches any annotation or marker} \\ {\tt .} & {matches a deletion made during this \WAVE{} session} \\ \end{tabular*} More complex search criteria can be defined by clicking on \button{More Options...} in the {\sf Find} window. This opens the {\sf Search Template} window (figure~\ref{fig:search-template}). \index{search template window@{\sf Search Template} window} \begin{figure} \centerline{\epsfig{file=search-template}} \caption{The {\sf Search Template} window.} \label{fig:search-template} \end{figure} In the {\sf Search Template}, you may select any combination of annotation fields as search criteria using the abbreviated menu buttons to the left of each field. Enter the values to be matched next to each active match control. Any fields marked by `{\sf Ignore}' are not used as match criteria. In figure~\ref{fig:search-template}, the {\sf Type}, {\sf Subtype}, and {\sf `Chan' field} criteria must be satisfied by an annotation in order for \WAVE{} to consider that annotation as a match; the {\sf Text} and {\sf `Num' field} criteria are ignored by \WAVE{}. If the {\sf Text} criterion is active, \WAVE{} requires that the specified text string appear in the {\tt aux} field of any matching annotation. Note that, as for search criteria specified in the {\sf Search for} field of the {\sf Find} window, the {\tt aux} field of a matching annotation is required only to \emph{contain} a matching text string, and additional text may appear in the {\tt aux} field before or after the matching string. As a shortcut, click on \button{Match selected annotation} to copy all fields of the selected annotation into the {\sf Search Template}. This action also sets all of the criteria menu buttons to {\sf Match}. \section{Saving your work} In a lengthy editing session, \WAVE{} saves your annotation file periodically. \index{annotation file!checkpointing} \begin{wrapfigure}[9]{r}{2.2cm} \mbox{\epsfig{file=file-save}} \end{wrapfigure} You can force \WAVE{} to save your edits \index{saving edits} at any time by selecting the {\sf Save} entry from the {\sf File} menu. Once all of the changes have been saved, \WAVE{} removes the parentheses from the title bar. \index{saving edits} When you are finished, exit from \WAVE{} as in the first exercise. It is not necessary to take any explicit action to save your edits, since \WAVE{} does so automatically on any normal exit (via \button{Quit} \index{Quit button@{\sf Quit} button} on the main control panel, or from the window menu). You will find two new files in the current directory. The file called `{\tt 100s.qrs}' is the edited version of the annotation file that you have just created, and `\verb+100s.qrs~+' is the \index{backup}\index{annotation file!backup} previous version (in this case, containing the unedited annotations as produced by {\tt sqrs}). At most one previous version is saved as a backup; if the original annotation file is not located in the current directory (i.e., if it was found elsewhere in the database path), \index{database path} it is left unchanged and \WAVE{} does not create a separate backup file. \section{Creating an annotation file manually} \index{annotation file!creating!manually} \index{creating an annotation file!manually} In this exercise, we created the annotation file using {\tt sqrs}, and then edited it. It is also possible to create an annotation file manually; to do so, simply fill in the {\sf Annotator} field \index{Annotator@{\sf Annotator}} in the {\sf Load} window \index{Load window@{\sf Load} window} \index{annotator name} with the annotator name of your choice. If there is no existing annotation file associated with the specified annotator for the current record, \WAVE{} creates an initially empty annotation file, to which you can add and edit annotations just as we did in this exercise. \section{Multi-edit mode} \label{sec:multi-edit} \begin{htmlonly} \index{multi-edit mode} \end{htmlonly} \begin{latexonly} \index{multi-edit mode|textit} \end{latexonly} \index{annotation!attached to signal} \index{signal!annotating independently} If you plan to use \WAVE{} with records containing signals that must be annotated independently, read this section. \index{View window@{\sf View} window} \index{chan in annotation@{\tt chan} in annotation} \index{annotation!chan@{\tt chan}} \index{signal!number} Using the \amenubutton{Show annotations} menu button in \WAVE{}'s {\sf View} window, \begin{wrapfigure}[6]{l}{6cm} \mbox{\epsfig{file=attach-to-signals}} \end{wrapfigure} you may choose to display annotations attached to signals, rather than in the center of the signal window as is the usual default. In this case, the {\tt chan} field of each annotation specifies the signal number of the signal to which it is attached. If annotation editing is enabled when annotations are displayed in this way, \WAVE{} is in \emph{multi-edit mode}. \index{signal!suppressing display} \index{signal!list} \WAVE{} draws the signals in order of signal number from the top to the bottom of the signal window, beginning with signal number 0 at the top of the signal window. When annotations are attached to signals, \WAVE{} draws them about 2 mm above the center of the range of the attached signal (except for special annotations that are displaced above or below the usual level). Although this presentation usually helps to avoid the visual confusion that might result from drawing annotations directly on the associated signals, it may contribute to confusion if the spacing between signals is too small (as may happen if the signal window is reduced in height, or if many signals are displayed). If this becomes a problem, try increasing the height of the signal window (by dragging on the resize handles on the window frame), or displaying only a subset of the signals (by specifying which signals are to be shown in the {\sf Signal list} field in \WAVE{}'s {\sf Analyze} window, and selecting `{\sf listed signals only}' from the \amenubutton{Draw:} menu in WAVE's {\sf View} window). In multi-edit mode, annotation editing operations are slightly different from those described above. The most important difference is that you must always point to the desired signal when inserting or moving annotations. In this mode, the {\tt chan} field in the {\sf Annotation Template} window does not determine the {\tt chan} field of an inserted annotation; rather, the signal to which you point determines the {\tt chan} field, and the {\tt chan} field in the {\sf Annotation Template} window is updated accordingly after each insertion. To move an annotation to a different signal, simply select it and drag the pointer to the desired time and signal. If you wish to change the {\tt chan} field of an annotation without changing its time, select the annotation and use the \keycap{$\uparrow$} and \keycap{$\downarrow$} keys to move it to the desired signal. Hold down \keycap{Ctrl} during these operations to copy the annotation, rather than to move it (simultaneous annotations are permitted if their {\tt chan} fields differ). \section{About link annotations} \label{sec:link-annotations} \index{link annotation}\index{annotation!link} A link (`{\tt @}') annotation permits external data to be associated with a specific time (and, optionally, a specific signal) in a record. To make link annotations visually distinctive, WAVE{} displays them in the color used for signals rather than that used for other annotations, and it underlines them. The {\tt aux} field of a link annotation contains a URL (uniform resource locator) that points to the external data, and an optional description displayed by \WAVE{} at the location of the annotation. (If the description is missing, \WAVE{} displays the URL name.) To view the external data associated with a link annotation, select it and then press \keycap{Return} (or \keycap{Enter}). \WAVE{} then directs your web browser to display the data specified by the URL. (If your web browser is not running, \WAVE{} starts it.) By default, \WAVE{} uses Mozilla. To configure \WAVE{} to use a different browser, see \hyperref{\WAVE{} and the Web} {section~} {, page~\pageref{sec:web}} {sec:web}. \index{Mozilla}\index{web browser} If the URL string does not contain a {\it protocol}{\tt ://} prefix (where {\it protocol} is typically {\tt http} or {\tt ftp}), \WAVE{} assumes that the URL refers to a file located somewhere in your WFDB path. In this case, \WAVE{} finds the file if possible, and attaches the necessary path information to the beginning of the URL before directing your web browser to display the associated data. \WAVE{} does not include built-in facilities for editing external data. You may insert and edit link annotations themselves, however, using the normal annotation editing facilities of \WAVE{}. To create a link annotation, select `{\tt @}' from the {\sf Type} menu of the {\sf Annotation Template}, and enter the URL in the {\sf Text} ({\tt aux}) field. If you wish to supply a description to be displayed by \WAVE{} in place of the URL, enter it after the URL, separating it from the URL with a space character. The current URL (i.e., the one associated with the most recently selected link annotation) can be passed by \WAVE{} to an external program such as a web browser or an HTML editor via the {\tt \$URL} menu variable (see the comments in \WAVE{}'s menu file for details). \section{Creating and using new annotation types} \label{sec:creating-annotation-types} \index{ANNTAB environment variable}\index{environment variables!ANNTAB} \index{X11 resources!{\tt Wave.Anntab}} \index{defining annotation types}\index{Type menu@{\sf Type} menu} \index{annotation!table of definitions} \WAVE{} was originally developed for use with ECGs, and the standard set of annotation types reflects this history. If you work with other types of signals, or if the standard set of annotations does not meet your needs, you may define your own annotation types and add them to the \amenubutton{Type:} menu in the {\sf Annotation Template}. To do so, first create a text file containing one-line entries defining each new type, as described in \hyperref{X11 Resources} {section~} {, page~\pageref{anntab},} {sec:resources} or see the file `{\tt anntab}' included in the \WAVE{} distribution for an example. Next, set the environment variable {\tt ANNTAB}, or the X11 resource {\tt Wave.Anntab}, to the name of this file. The next time you run \WAVE{}, the new types from your {\tt ANNTAB} file will appear in the \amenubutton{Type:} menu. You may use them in the same way as the standard annotation types. Any annotation files you create or edit while your {\tt ANNTAB} file is loaded will contain embedded copies of your type definitions, so that these files can be used on other systems and by other users without the need to reference your {\tt ANNTAB} file each time. \index{annotation!type codes} \index{unassigned annotation type codes} In most cases, you should use the unassigned type codes between 42 and 49 for custom annotation types, rather than redefining the standard types. If you are not working with ECGs, however, you may redefine any type codes you wish (the legal range is defined in `{\tt '}). \chapter{Analyzing Data with \WAVE{}} For many users, \WAVE{}'s data analysis capabilities will be of greatest interest. \WAVE{} itself does not perform analysis of the signals or annotations it displays. What \WAVE{} offers is an easy-to-use means of controlling external analysis programs and viewing their outputs. The buttons in the {\sf Analyze} window activate these external programs. These buttons are easy to create -- in fact, you can add your own buttons to those in the {\sf Analyze} window while \WAVE{} is running. In the previous chapter, we used the \button{Mark QRS complexes} button to run the standard WFDB application program {\tt sqrs}. In this chapter, we will see how \WAVE{} controls {\tt sqrs}, and we will create and test a new button to generate a heart rate signal using {\tt sqrs} and another standard WFDB application, {\tt tach}. Discussions of \WAVE{}'s ``logbook'' and ``oscilloscope'' facilities follow, and then we study an extended example of how to develop an analysis program controlled by \WAVE{}, using the C programming language. At the end of this chapter, we explore how this relationship can be inverted, so that \WAVE{} can be driven by an external program. As before, use \WAVE{} to open record {\tt 100s}: \begin{verbatim} wave -r 100s \end{verbatim} Click right on \menubutton{File}, click left on {\sf Analyze...}, and once again, click left on \button{Mark QRS complexes}. \index{analysis commands window@{\sf Analysis Commands} window} Where does the command that appears in the {\sf Analysis Commands} window, \begin{verbatim} sqrs -r 100s -f 0 -t 1:00.000 -s 0 \end{verbatim} come from, and what does it mean? \index{sqrs command@{\tt sqrs} command} If you refer to the {\tt man} page for {\tt sqrs}, \htmladdnormallink{{\tt sqrs(1)}}{../wag/sqrs-1.htm} (either by typing `{\tt man sqrs}' in a terminal emulator window, or by looking it up in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}), you will see that the second command-line argument ({\tt 100s}) is (not surprisingly) the record name, `{\tt -f 0}' instructs {\tt sqrs} to start at the beginning of the record, `{\tt -t 1:00.000}' tells it to stop 1 minute later, and `{\tt -s 0}' specifies that signal 0 is to be used for QRS detection. \section{\WAVE{}'s menu file} \label{sec:menu-file} \index{menu file} To understand where the command comes from, we need to inspect \WAVE{}'s \emph{menu file}. \WAVE{} allows you to edit its menu file using any editor of your choice. By default, \WAVE{} uses a simple Open Look text editor called {\tt textedit}, which is included with the XView package used by \WAVE{}. If you prefer to use another editor such as {\tt emacs}, set and export the {\tt EDITOR} environment variable before starting \WAVE{} (you may wish to include this step in your {\tt .login} or {\tt .cshrc}). Most other UNIX applications that invoke an external text editor use the {\tt EDITOR} environment variable in the same way. \index{EDITOR environment variable} \index{environment variables!EDITOR} \index{textedit command@{\tt textedit} command} Click left on the \button{Edit menu} button in the {\sf Analyze} window. \begin{wrapfigure}[12]{r}{7.25cm} \mbox{\epsfig{file=system-menu}} \end{wrapfigure} Unless your \WAVE{} menu has already been customized, \WAVE{} pops up a notice as shown \begin{latexonly} at right. \end{latexonly} \begin{htmlonly} above. \end{htmlonly} Click left on \button{Copy} to continue with this exercise. After a few seconds, the \WAVE{} menu file appears (in an OpenWindows {\tt textedit} window as in figure~\ref{fig:wave-menu}, unless you have specified a different editor by setting the {\tt EDITOR} environment variable). \index{EDITOR environment variable} \index{environment variables!EDITOR} \index{textedit command@{\tt textedit} command} \begin{figure} \centerline{\epsfig{file=wave-menu}} \caption[\WAVE{}'s menu file]{The beginning of \WAVE{}'s menu file, in a {\tt textedit} window.} \label{fig:wave-menu} \end{figure} The comments at the beginning of the \WAVE{} menu file describe how it works. Each of the buttons in the {\sf Analyze} window (except for those in the first three rows) corresponds to an entry in the menu file. The first part of the entry -- up to the first tab character -- specifies the label that appears on the button; the remainder of the entry specifies the command that is inserted into the {\sf Analysis Commands} window if you click left on the button. As noted in the comments, certain strings that begin with `{\tt \$}' are recognized as \index{menu variables}\index{variables!in \WAVE{} menu file} \emph{menu variables} and are interpreted by \WAVE{} before being passed to the shell in the {\sf Analysis Commands} window. In the case of the {\tt sqrs} command that appears in the {\sf Mark QRS complexes} entry, these strings are `{\tt \$RECORD}', `{\tt \$START}', `{\tt \$END}', and `{\tt \$SIGNAL}'. \WAVE{} replaces `{\tt \$RECORD}' with the record name, as it appears in the {\tt Record} field of the {\sf Load} window. `{\tt \$START}' is the time of the `{\tt <}' marker, shown in the {\sf Start} field of the {\sf Analyze} window; if no `{\tt <}' marker has been defined, the {\sf Start} field contains the string `{\sf beginning of record}', and `{\tt \$START}' becomes {\tt 0}. Similarly, `{\tt \$END}' is the time of the `{\tt >}' marker, shown in the {\sf End} field; if no `{\tt >}' marker has been defined, the {\sf End} field is `{\sf end of record}', and `{\tt \$END}' is determined by the record length as encoded in the header file for the record (see \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}). Finally, `{\tt \$SIGNAL}' is taken from the {\sf Signal} field of the {\sf Analyze} window; it specifies which signal is to be analyzed. (For applications that can analyze more than one signal, the string `{\tt \$SIGNALS}' is taken from the {\sf Signal list} field.) Examine the commands defined in the default \WAVE{} menu; they should give you an idea of what is possible. \index{Analyze window@{\sf Analyze} window} \index{analysis commands window@{\sf Analysis Commands} window} \index{menu variables} \index{signal!list} \section{Defining a region of interest} \index{region of interest} \index{marker!{\tt >}} \index{marker!{\tt <}} A common problem in many cases is simply selecting a segment to be analyzed, from a digitized signal file that may include extraneous material at the beginning or at the end. As you can see from the {\tt sqrs} example, this can be done easily by inspecting the signals with \WAVE{} and placing `{\tt <}' and `{\tt >}' markers at the beginning and end of the region of interest. If the analysis program is written to accept starting and ending times (more on this subject below), these times can be passed as command-line arguments by \WAVE{}. Similarly, it's simple to choose one or more of a number of signals to analyze after viewing them with \WAVE{} (either type into the {\sf Signal} field directly, use the increment/decrement buttons, or point to a signal in the signal window, press and hold the \keycap{Shift} key, and click left). Close the {\tt textedit} window and experiment with changing the {\sf Start}, {\sf End}, and {\sf Signal} fields of the {\sf Analyze} window, and running {\tt sqrs}, until you are familiar with the results. \label{sec:analysis-example} \section{Analysis example: Generating a heart rate signal} In the rest of this exercise, we'll develop a command for generating a heart rate signal, add it to the \WAVE{} menu, and test it. First, set the {\sf Start} and {\sf End} fields to {\tt 0} and {\tt 1:0} respectively, and click on \button{Mark QRS complexes} to regenerate the annotation file. Open the annotation file using the {\sf Load} window. \index{WFDB Software Package} \index{tach command@{\tt tach} command} Generating a heart rate signal is simple enough, given a beat annotation file and \htmladdnormallink{{\tt tach}}{../wag/tach-1.htm}, a program supplied as part of the WFDB Software Package. {\tt tach} reads an annotation file, calculates an instantaneous heart rate for each cardiac cycle from the reciprocal of the interval between successive beat annotations, and then uniformily resamples this signal (by default, at 2 Hz). Referring to the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, we find the command we need to generate a heart rate signal from `{\tt 100s.qrs'}: \begin{verbatim} tach -r 100s -a qrs -f 0 -t 1:0 \end{verbatim} \index{analysis commands window@{\sf Analysis Commands} window} If this command is typed into the {\sf Analysis Commands} window, it prints: \begin{verbatim} 73.9726 73.9726 73.7238 . . . 75.7788 74.5703 73.8015 \end{verbatim} These numbers are the calculated samples of the heart rate signal (the units are beats per minute). We can generalize this command as: \begin{verbatim} tach -r $RECORD -a $ANNOTATOR -f $START -t $END \end{verbatim} where the parameters {\tt \$RECORD}, {\tt \$ANNOTATOR}, etc., are replaced by the appropriate values when the command is executed. This is the form of the command we should add to \WAVE{}'s menu; by parameterizing the arguments in this way, we can use \WAVE{} to select a record, annotator, and segment to be analyzed as for the other commands in the menu. To add this command to the menu, click on \button{Edit menu} again. Position the cursor at the end of the menu file and add the line: \begin{verbatim} Heart ratetach -r $RECORD -a $ANNOTATOR -f $START -t $END \end{verbatim} where {\tt } represents one or more {\sf TAB} characters; at least one {\sf TAB} must be present (spaces don't count). Be sure to end the line with {\sf RETURN}. Save the menu (choose {\sf Save} from the {\tt textedit} window's \menubutton{File} menu), and quit from {\tt textedit}. Now \index{textedit command@{\tt textedit} command} click on \button{Reread menu} in the {\sf Analyze} window. The {\sf Analyze} window will disappear briefly, then reappear (possibly in a different location, depending on your window manager) with a new \button{Heart rate} button. If you click on this button, \WAVE{} executes the {\tt tach} command as above. You might prefer to look at heart rate as a signal rather than as a column of numbers. {\tt tach} has the capability of generating an output signal file instead of text. To try this out, edit the menu file again, and add `{\tt -o hr\_\$RECORD}' to the end of the entry. (You may split the command across two lines using a `{$\tt \backslash$}' at the end of the first line if you wish.) Once again, save the menu file, quit from {\tt textedit}, click on \button{Reread menu}, and on \button{Heart rate}. This time, there is no text output from {\tt tach}; if you enter `{\tt hr\_100s}' in the {\sf Record} field of the {\sf Load} window, however, \WAVE{} displays the calculated heart rate signal. \begin{figure} \centerline{\epsfig{file=main-with-hr}} \caption{Heart rate signal.} \label{fig:main-with-hr} \end{figure} \index{time scale}\index{amplitude scale} \index{scales!setting}\index{display scale} \index{View window@{\sf View} window} (You may wish to change the time and amplitude scales using the {\sf View} window, in order to see more of the signal at once. If you set the time scale to 250 mm/min and the amplitude scale to 100 mm/mV, you will be able to see the heart rate signal for the entire record as in figure~\ref{fig:main-with-hr}; the periodic modulation is due to respiratory sinus arrhythmia.) \section{The signal list} \index{signal!list} Although this exercise used only one signal of the input record, many applications will use two or more signals simultaneously. The {\sf Signal list} initially contains the signal numbers of all available signals; as noted above, the signal list can be passed to an application listed in the menu file as a set of command-line arguments in place of the string `{\tt \$SIGNALS}'. Although you can type into the {\sf Signal list} field directly to change it, you can also do so using the mouse. Pointing to a signal, hold down the \keycap{Control} key while clicking left to add it to the signal list, or hold down the \keycap{Meta} key while clicking left to delete its first occurrence from the signal list. (Most keyboards do not have a key labelled {\sf Meta}. On most Sun keyboards, the {\sf Meta} keys look like this: \keycap{$\Diamond$}; they flank the space bar. On most other keyboards, the \keycap{Alt} key is equivalent to {\sf Meta}.) \index{Meta key@{\sf Meta} key} \section{Using a customized menu file} \index{WAVEMENU environment variable}\index{environment variables!WAVEMENU} \index{menu file} \index{default menu} Recall that when you first examined the menu file, \WAVE{} warned you to set the {\tt WAVEMENU} variable the next time you run \WAVE{}. {\tt WAVEMENU} should be set to the name of your customized menu file (you can give the menu file any name you choose; {\tt wavemenu} is the default name). For example, if you use the C-shell and {\tt wavemenu} is in your home directory, use the command \begin{verbatim} setenv WAVEMENU ~/wavemenu \end{verbatim} before running \WAVE{} again. You may wish to add this command to your \index{login@{\tt .login}} {\tt .login} script to set the variable automatically each time you log in. (If you forget to set {\tt WAVEMENU}, \WAVE{} will find your customized menu file only if it is in the current directory and is named {\tt wavemenu}; otherwise, \WAVE{} will use the system default menu again.) \section{Using the {\sf Log} window} When examining a recording, you may find features that you wish to document for further study. In addition to its capabilities for interactive annotation editing, \WAVE{} also offers an easy-to-use ``logbook'' for collecting brief notes about segments to which you may wish to return. Each entry in a \WAVE{} log file contains the record name, time, and an optional comment -- so it is possible to put together a collection of examples taken from any number of records in a single log. Once the log has been created, the entries in it can be reviewed very easily, or the log can be used as a script for {\tt pschart}, which will print the signals, annotations, and your comments for each entry, changing records automatically. Open the {\sf Log} window (see figure~\ref{fig:log-window}) by choosing {\sf Log...} from \WAVE{}'s \menubutton{File} menu. \begin{figure} \begin{htmlonly} \index{Log window@{\sf Log} window} \end{htmlonly} \begin{latexonly} \index{Log window@{\sf Log} window|emph} \end{latexonly} \centerline{\epsfig{file=log-window}} \caption{The {\sf Log} window.} \label{fig:log-window} \end{figure} When you first open the {\sf Log} window, many of its controls are dimmed until you specify a log file and begin adding entries to it. To create a log file, or to review or add to an existing one, enter its name in the {\sf File:} field of the {\sf Log} window. To add an entry, use \WAVE{}'s navigation controls so that the segment upon which you wish to comment appears in the signal window. Enter your comments in the {\sf Description:} field, then click on \button{Add}. The log navigation buttons become visible once there are log entries recorded. You can use \button{\tt <} and \button{\tt >} to step through the log one entry at a time, or use \button{\tt <<} or \button{\tt >>} to present a ``slide show'' (control the length of time each entry appears using the {\sf Delay:} slider, and use \button{Pause} to interrupt the show). \button{\tt |<} and \button{\tt >|} can be used to jump directly to the beginning or the end of the log. To replace a log entry, first use the log navigation buttons to display that entry. Make any adjustments necessary (use the navigation buttons in \WAVE{}'s main window to reposition the signal window as needed, and correct the log entry text in the {\sf Description:} field), then click on \button{Replace}. To delete a log entry, select it first as above, then click on \button{Delete}. For more complex editing (such as rearranging entries), click on \button{Edit} to open the log file in a {\tt textedit} window (or in the editor of your choice, if you have set the {\tt EDITOR} environment variable). After completing your edits, save them and exit from the editor, then click on \button{Load} to reload the edited log into \WAVE{}. (You can also use \button{Load} if another process, such as one started from the {\sf Analyze} window, has modified the log file.) \label{sec:printing-log} \index{printing!log with charts} \index{log!printing with charts} \WAVE{}'s default menu file contains a commented-out entry with the tag ``Print log with charts''. If a button with this text does not appear in the {\sf Analyze} window, click on \button{Edit menu}, remove the comment characters from this entry, save the menu and exit the editor, and click on \button{Reread menu}. You can now print the log together with the signals and annotations by clicking on \button{Print log with charts}. \newpage \section{Using the {\sf Scope} window} \index{Scope window@{\sf Scope} window} \begin{wrapfigure}{l}{2cm} \mbox{\epsfig{file=scope-window}} \caption[The {\sf Scope} window.]{\sf Scope} \end{wrapfigure} Some types of signals are characterized by changes in waveform shape (morphology), which may occur gradually or suddenly. For example, ischemic ST changes in the ECG are usually visible as quasi-continuous changes over a period of at least 30 seconds (often much longer). It can be helpful to view such signals on an oscilloscope, where the trigger can be adjusted so that the (approximately) periodic waveforms are superimposed. Such displays, and their digital counterparts, are very widely used in analysis of long-term ECG recordings. \WAVE{}'s {\sf Scope} window offers some of the features of this type of display, with a few unique capabilities and limitations of its own. To open the {\sf Scope} window, first open the {\sf Analyze} window, then click left on \button{Show scope window}. The {\sf Scope} window simulates an oscilloscope by using color map animation (on hardware with an X server that supports this technique) or stipple masks (otherwise). The display is ``triggered'' by beat (QRS) annotations (other annotations, such as rhythm changes, are ignored). As each waveform is drawn, the previously drawn waveform is made to fade progressively into the background, disappearing completely after four new waveforms have been drawn. The appearance is similar to that of an analog scope with moderately long-persistence phosphor. The x-position in the window of the trigger point (the sample that the annotation points to for each waveform) is adjustable using the {\sf dt:} setting (the value is expressed in seconds, and may be negative). It is also possible to resize the window, so that several waveforms can be viewed side-by-side as well as superimposed. As shown in the figure at left, \WAVE{} shifts waveforms to the bottom of the window if they have been annotated as ventricular ectopic beats. (This behavior was chosen to be most useful for studying ECGs, but other signals can be annotated in the same way if desired.) The navigation controls are similar to those in the {\sf Log} window. \button{\tt <} and \button{\tt >} may be used to move backward and forward through the record one beat at a time. \button{\tt <<} and \button{\tt >>} provide a continuous ``movie'' (at a speed controlled by the {\sf Speed} slider) that can be interrupted by \button{Pause}, which also recenters the signal window on the current beat. Whenever the current beat is visible in the signal window, it is marked with the selection rectangle. \index{signal!selecting} Only one signal at a time can be displayed in the {\sf Scope} window. Select that signal using the {\sf Signal} field in the {\sf Analyze} window, or press and hold the \keycap{Shift} key while clicking left in the signal window near the desired signal. \section{Analysis example: Measuring ST deviations} We now turn to an extended example showing how an application can be written in C to be used with \WAVE{}. The application is an example only; it is not intended for any serious application, clinical or otherwise. Its structure illustrates, however, how a real application can be built. As noted in the previous section, changes in the ST segment of the ECG are of interest because of their relationship to ischemia (which occurs when the oxygen supplied by the coronary arteries is insufficient to meet the demands of the myocardium). Characteristically, ischemia affects the shape of the ST segment, which corresponds to the first phase of ventricular repolarization following the QRS complex; typically, the ECG signal may not return to its baseline, or isoelectric, level (defined in the interval between the P-wave and the QRS complex) until after the T-wave. The level of the ST segment relative to the isoelectric level is the ST deviation. Conventionally, the ST deviation is measured at a single point 80 milliseconds after the J point (the end of the QRS complex). An ST deviation of 100 $\mu$V is considered to be clinically significant, and consistent with ischemia. We will design a program for measuring ST deviations using \WAVE{}. We assume that the QRS complexes have been marked, and that, in addition, two special annotations (`{\tt (}', signifying the beginning of the QRS complex, and `{\tt )}', signifying its end) have also been marked for the first beat that we wish to measure. (To use our program on an unannotated record, we can use \button{Mark QRS complexes}, and then insert the special annotations manually around the first beat.) Let us begin by deciding how the program should be invoked from \WAVE{}. At a minimum, we should be able to specify which signal should be measured, and the region of interest. \index{record!name} \index{annotator name} The program also needs to know the record and annotator names (the latter so that it can read the annotations it needs to get started). Following the conventions used by many other WFDB applications, we add the following entry to \WAVE{}'s menu file: \begin{verbatim} Measure ST deviationstdev -r $RECORD -a $ANNOTATOR \ -f $START -t $END -s $SIGNAL \end{verbatim} %$ % <-- this is here to keep emacs happy in fontlock mode \htmladdnormallink{Here}{stdev.c} is the C source for the program. If this looks unfamiliar, read about how to write WFDB applications in the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm}. \begin{latexonly} {\small \verbatiminput{../examples/stdev.c} } \end{latexonly} The source for this program is included in the \WAVE{} distribution as {\tt stdev.c}. Copy this file to the \WAVE{} host if necessary, and compile it in a terminal window using a command such as \begin{verbatim} cc -o stdev -O stdev.c -lwfdb \end{verbatim} Depending on how the WFDB Software Package has been installed on the \WAVE{} host, you may need to use {\tt -I/usr/local/include} and {\tt -L/usr/local/lib} options so that the compiler can locate the {\tt *.h} files and the WFDB library. On some systems, the C compiler may have a different name, such as {\tt gcc}. Consult an expert such as your system administrator if in doubt about compiling a C program with the WFDB library. Once the source has been compiled successfully, install the executable binary file {\tt stdev} in a directory in your {\tt PATH}. If you are using the C-shell, you may need to run the command {\tt rehash} to make your shell aware that the program has been installed. Test the program in the terminal window by typing {\tt stdev} \noindent which should produce a brief summary of its options. \emph{If this test fails, figure out why and correct the problem before continuing.} Once the program has been compiled and installed successfully, try running it from within \WAVE{}. Remember that the annotation file must include the `{\tt (}' and `{\tt )}' annotations to mark the beginning and end of the first beat to be measured. Using it on signal 0 of record 100s, with a suitably augmented set of reference annotations, I obtained output beginning with: \begin{verbatim} 0.0171296 -84 0.0306481 -24 0.0437963 -9 0.0569907 -34 0.0701389 -64 0.08375 -59 0.0946296 -49 0.111204 -29 0.125278 -4 0.138796 -59 0.151944 -9 0.164815 -34 \end{verbatim} (Your output will vary, depending on the exact locations of the `{\tt (}' and `{\tt )}' annotations that you insert.) \index{plt command@{\tt plt} command} If {\tt plt} has been installed on your system, you can use it to plot this output (see \htmladdnormallink{{\tt plt(1)}}{../wag/plt-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details on {\tt plt}). To do so, modify the entry in the menu file, so that it becomes: \begin{verbatim} Plot ST deviationstdev -r $RECORD -a $ANNOTATOR \ -f $START -t $END -s $SIGNAL | \ plt 0 1 \ -t "ST deviations, Record $RECORD" \ -x "Elapsed time (minutes)" \ -y "ST deviation (microvolts)" \end{verbatim} If you now reread the menu and use the new \button{Plot ST deviation} button, {\tt plot2d} opens a window with a plot in it, similar to figure~\ref{fig:stdev-plot}. \begin{figure} \centerline{\epsfig{file=stdev}} \caption{ST deviations measured by the example program.} \label{fig:stdev-plot} \end{figure} (To make this plot, I used the first ten minutes of record mitdb/100; the sample record 100s contains only the first minute of data shown in the plot.) This window stays open until you dismiss it by pressing \index{analysis commands window@{\sf Analysis Commands} window} \keycap{Enter} (\keycap{Return}) in the {\sf Analysis Commands} window; if you forget to do so, the next command you run from the {\sf Analyze} window will serve this function (but the command itself won't run, since {\tt plot2d}, rather than the command interpreter, will have read the command string). \subsection*{Further work} If you would like to make this program work better, here are a few ideas: \begin{itemize} \item Extending it to make measurements on multiple signals is an easy project. Allow the user to set measurement reference points on each signal separately. \item It's also not too difficult to copy the input annotation file, writing the measurements into the {\tt num} fields of the annotations, so that \WAVE{} can display them as a signal. If you try this, keep in mind that {\tt num} values must be integers in the -128 to 127 range, so don't record the deviations in $\mu$V. If you want to write the measurements into a \emph{signal} file (as we did with {\tt tach} earlier in this chapter), this can get complicated---see the source for {\tt tach} for hints about doing this. \item The measurement point should move closer to the J point at high heart rates; see any reference on the ECG for a discussion of this issue. \item The measurements as made above are very easily contaminated by small amounts of noise in the signals. You should be able to improve them significantly by averaging the amplitudes of several samples in the neighborhoods of the isoelectric point and the ST level measurement point. \item The program should not produce measurements for ventricular ectopic beats, or for beats that are extremely noisy. Noise detection is an interesting research problem in itself. \item You may wish to examine using beat averaging (or median filtering) to reduce noise in the measurements further; the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm} describes a beat averager that may be a useful starting point. \item Another source of measurement error is baseline wander in the ECG, visible as slopes in the intervals between beats. You can reduce this error by \emph{modelling} the baseline drift and subtracting it from the signal. Simple linear models, or quadratic or cubic splines, can be used. \end{itemize} \section{Controlling \WAVE{} from an external program} \label{sec:wave-remote} Although the facilities discussed in the previous sections provide highly flexible control of external analysis programs, in some cases it is desirable for another program to control \WAVE{} rather than for that program to run under \WAVE{}'s control. The program {\tt wave-remote}, provided with the \WAVE{} distribution, can drive \WAVE{}'s display. Using {\tt wave-remote}, you can cause \WAVE{} to go to any specified location in the current record, to load another set of annotations for the current record, or to open another record. By providing a command-line interface for driving \WAVE{}, {\tt wave-remote} simplifies the task of driving \WAVE{} from other programs or shell scripts; in general, it is sufficient merely to fork a new process and invoke {\tt wave-remote} with the appropriate arguments (for example, using the {\tt system()} function provided in the standard C library). If it is not acceptable to start a new process for this purpose, {\tt wave-remote} is also provided in C source form, and the functions it uses to drive \WAVE{} can be invoked directly from your program. Options for {\tt wave-remote} are: \begin{description} \item[{\tt -pid} \textit{processid}] Control the \WAVE{} process with the specified \textit{processid}. (If this option is omitted, and more than one instance of \WAVE{} is running, {\tt wave-remote} controls the one with the highest \textit{processid}.) \item[{\tt -r} \textit{record}] (Re)open the specified \textit{record}. \item[{\tt -a} \textit{annotator}] (Re)open the specified \textit{annotator} for the current record. \item[{\tt -f} \textit{time}] Go to the specified \textit{time} in the current record. \item[{\tt -s} \textit{signal ...}] Display the specified \textit{signal(s)} only. (Use signal numbers as in the {\sf Signal list} in the {\sf Analyze} window.) \end{description} For example, the command \begin{verbatim} wave-remote -r 100s -a atr -f 0:22 \end{verbatim} \noindent causes \WAVE{} to open record {\tt 100s}, with annotator {\tt atr}, and to show data starting at 22 seconds after the beginning of the record. If \WAVE{} is not running when {\tt wave-remote} is invoked, {\tt wave-remote} launches \WAVE{} provided that the record to be opened has been specified. {\tt wave-remote} exits immediately once it has delivered its instructions to \WAVE{}. As an alternative to the command-line interface offered by {\tt wave-remote}, the {\tt wavescript} application provides the same services, but is controlled by a script file (named on the {\tt wavescript} command line). Scripts for {\tt wavescript} should contain on each non-comment line a single option/argument pair as described above for {\tt wave-remote}. In addition to {\tt wave-remote}'s options, {\tt wavescript} also accepts a {\tt -p {\em path}} option; the {\tt \em path} argument is appended to the WFDB path. Any line in the script that does not begin with a `{\tt -}' is treated as a comment line. For example, if the file {\tt example.xws} contains \begin{verbatim} # Here is a comment -r 100s -a atr -f 0:22 \end{verbatim} \noindent then `{\tt wavescript example.xws}' has the same effect as the {\tt wave-remote} example above. Using these interfaces, an analysis program (or any program, such as a web browser) can drive \WAVE{}'s display automatically. In a typical application, a semi-automated analysis program might require the user to make or correct annotations for portions of a record that are selected as part of the program's analysis (and that cannot be known \emph{a priori}). An analysis program that runs {\tt wave-remote} or {\tt wavescript} may itself be running within \WAVE{}'s own {\sf Analysis Commands} window. This might be useful to permit the user to set the region of interest interactively, before the analysis begins and takes over control of \WAVE{} to show its progress or to obtain additional input. \section{\WAVE{} and the Web} \label{sec:web} In this section we explore how World Wide Web documents and WFDB records (i.e., signals and annotations viewable using \WAVE{}) can be linked using \WAVE{} together with a web browser. Beginning with \WAVE{} 6.0, it is possible while running a web browser to click on a link from a web document to a specified location in a WFDB record and have \WAVE{} open the record at that location. It is similarly possible while running \WAVE{} to click on a \hyperref{link annotation} {link annotation (see section~} {, page~\pageref{sec:link-annotations})} {sec:link-annotations} and have your web browser open the data specified by the link. \subsection*{Controlling \WAVE{} from a web browser} Most web browsers, such as Firefox or Mozilla, can use so-called \emph{helper} or \emph{viewer} applications. These are external applications that are invoked by the browser to present data in formats that (usually) are not supported by the browser directly. For example, browsers often use {\tt ghostscript} to display PostScript data. It is possible to configure Mozilla and other web browsers so that \WAVE{} can be used indirectly as a helper application. \index{Mozilla}\index{Firefox}\index{web browser} By using \hyperref{{\tt wavescript}} {{\tt wavescript} (see section~} {, page~\pageref{sec:wave-remote})} {sec:wave-remote} as the helper application for viewing WFDB records, an already-running \WAVE{} process can be made to open a record and an annotation set and to move to a specified location in the record. With this approach, it is not necessary to start a new \WAVE{} process each time. On most platforms, \WAVE{} can be started automatically if it is not running already; under MS-Windows, however, you must start \WAVE{} manually before using it as a browser helper application. You can set up your browser to view files with the MIME type {\tt appli\-cation/\-x-wavescript} using {\tt wavescript}. Some browsers, including Firefox, Mozilla, and Netscape, also allow you to specify that any URL with the {\tt .xws} suffix should be handled by {\tt wavescript}. If you use Firefox, click on any {\tt .xws} link \begin{rawhtml} (such as this one) \end{rawhtml} and Firefox will open a dialog box asking what it should do. \begin{figure} \centerline{\epsfig{file=firefox-new-helper}} \caption{Adding {\tt wavescript} as a Firefox helper: step 1.} \label{fig:firefox-new-helper} \end{figure} Choose ``Open with'', click on ``Browse...'', and choose {\tt wavescript} (usually in {\tt /usr/bin}) in the file selection dialog, then check ``Do this automatically for files like this from now on.'' The dialog box will expand to include instructions for modifying your choice. \begin{figure} \centerline{\epsfig{file=firefox-new-helper2}} \caption{Adding {\tt wavescript} as a Firefox helper: step 2.} \label{fig:firefox-new-helper} \end{figure} Dismiss the dialog by clicking ``OK'', and \WAVE{} should open the record specified by the {\tt .xws} file. \index{Firefox} If you use Mozilla, choose {\sf Preferences} from the {\sf Edit} menu, then open the {\sf Navigator} category and choose {\sf Helper Applications}. Click on {\sf New Type...} to pop up a dialog, where you should enter '{\tt WAVEScript}' as the {\sf Description of type}, '{\tt xws}' as the {\sf File extension}, '{\tt application/x-wavescript}' as the {\sf MIME type}, and '{\tt wavescript}' as the {\sf Application to use}' (omitting the quotation marks in each case). Click {\sf OK} to close the {\sf New Type} dialog, and click {\sf OK} in the {\sf Preferences} dialog to save your changes. \index{Mozilla} \begin{figure} \centerline{\epsfig{file=mozilla-new-helper}} \caption{Adding {\tt wavescript} as a Mozilla helper.} \label{fig:mozilla-new-helper} \end{figure} If your browser is Netscape 4.x, do this by choosing {\sf Preferences} from the {\sf Edit} menu, opening the {\sf Navigator} category in the {\sf Preferences} window, selecting {\sf Applications}, then clicking on the {\sf New...} button. Fill in the dialog box as shown in figure~\ref{fig:netscape-new-helper} (enter '{\tt WAVEScript}' as the {\sf Description}, '{\tt application/x-wavescript}' as the {\sf MIMEType}, '{\tt xws}' as the {\sf Suffixes}, click on the radio button next to {\sf Application} and enter '{\tt wavescript \%s}' (note the '{\tt \%s}') as the {\sf Application}, then save your settings by clicking on {\sf OK} in each of the dialogs. (Other versions of Netscape have similar dialog boxes.) \index{Netscape} \begin{figure} \centerline{\epsfig{file=netscape-new-helper}} \caption{Adding {\tt wavescript} as a Netscape helper.} \label{fig:netscape-new-helper} \end{figure} If you use Galeon, choose {\sf Preferences} from the {\sf Settings} menu, then select {\sf Handlers}, then {\sf MIME Types}, then click on {\sf New}. Enter '{\tt application/x-wavescript}' as the MIME type and '{\tt wavescript}' as the Helper (note that the quotation marks are not part of either entry). Select {\sf Run with helper} and {\sf Always use this helper}; do not select {\sf Run in terminal} or {\sf Processes URL}. Save your changes by clicking on {\sf Close}. Note that Galeon does not support using the file suffix to select a helper application, so Galeon will only be able to control \WAVE{} if the web server is properly configured to transmit the {\tt applica\-tion/\-x-wave\-script} MIME type for {\tt .xws} files (see \hyperref{Serving .xws files} {Serving .xws files, section~} {, page~\pageref{sec:serving-xws}} {sec:serving-xws}). \index{Galeon} \begin{figure} \centerline{\epsfig{file=galeon-new-helper}} \caption{Adding {\tt wavescript} as a Galeon helper.} \label{fig:galeon-new-helper} \end{figure} Similar operations may be possible with other browsers; consult your browser's documentation. \begin{latexonly} \begin{wrapfigure}[3]{l}{1.5cm} \mbox{\epsfig{file=wave-icon}} \end{wrapfigure} Once {\tt wavescript} has been installed as a helper application, your web browser will launch {\tt wavescript} whenever you click on a link to a {\tt .xws} script file. As an example, the on-line version of this guide contains a link to the {\tt example.xws} script shown in the previous section. \end{latexonly} \begin{htmlonly} Once {\tt wavescript} has been installed as a helper application, your web browser will launch {\tt wavescript} whenever you click on a link to a {\tt .xws} script file. As an example, here is a link to the {\tt example.xws} script shown in the previous section: \end{htmlonly} \begin{rawhtml} click here to view waveforms \end{rawhtml} To provide such a link in an HTML document, use a statement such as \begin{verbatim} click here to view waveforms \end{verbatim} \begin{latexonly} The {\tt wave.png} icon (shown above) is provided in the \WAVE{} distribution. \end{latexonly} \subsection*{Serving {\tt .xws} files} \label{sec:serving-xws} If you plan to make {\tt .xws} files available from a Web server, you will also need to configure your Web server so that it sends the appropriate MIME type encoding ({\tt application/x-wavescript}) whenever a browser requests a {\tt .xws} file. (When you use a browser to read a local file, this is not required, since no Web server is involved in this case; rather, the local {\tt .mime.types} file is used by the browser to recognize the type based on the file suffix.) To configure the Apache {\tt httpd} (at this time, the most widely-used Web server), find the file {\tt httpd.conf} (typically in {\tt /etc/httpd/conf}) and add the line \begin{verbatim} AddType application/x-wavescript .xws \end{verbatim} \noindent in the section that contains other {\tt AddType} directives. Restart the server to force it to reread {\tt srm.conf}. The NCSA {\tt httpd} and variants of it can be configured in the same way; if you use the CERN {\tt httpd}, the syntax of the {\tt AddType} directive is different: \begin{verbatim} AddType .xws application/x-wavescript text \end{verbatim} \noindent If you use some other Web server, consult its documentation to see how to proceed. Note that readers of Web pages with a link to a {\tt .xws} file will need to have \WAVE{} and {\tt wavescript}, as well as the WFDB record indicated in the {\tt .xws} file, available to them \emph{on the system on which their Web browser is running}. Note that if you attempt to read a {\tt .xws} file before completing the necessary configuration steps, your browser's cache may interfere with proper operation after the configuration is completed. If this happens, simply delete any {\tt .xws} files from your browser's cache. \subsection*{Controlling a web browser from \WAVE{}} \index{Galeon}\index{Konqueror}\index{Mozilla}\index{Netscape}\index{Opera} \index{web browser} On Unix and Unix-like platforms such as GNU/Linux, \WAVE{} uses a web browser to display external data associated with link annotations, and to display the on-line version of this guide. A suitable web browser must be installed on the \WAVE{} host in order for this to work. By default, \WAVE{} uses Firefox (\htmladdnormallink{{\tt http://www.mozilla.org/firefox/}}{http://www.mozilla.org/firefox/}); several other browsers, including Galeon, Konqueror, Mozilla, Netscape, and Opera, can also be controlled by \WAVE{}. If you prefer to use one of these rather than Firefox, set the environment variable {\tt URLV} to the name of your browser. \WAVE{} controls the web browser in the same way it controls other external programs: by running commands in the {\sf Analysis Commands} window. \WAVE{}'s interface to the web browser is defined by the line beginning with the {\tt } tag in \WAVE{}'s menu file. In the standard version of the menu file, this line reads: \begin{verbatim} url_view $URL \end{verbatim} % $ (this comment is here only to keep Emacs's LaTeX fontification happy) \noindent As this example illustrates, the menu variable {\tt \$URL} can be used to pass the selected URL from \WAVE{} to your browser. (\WAVE{} expands any incomplete URLs from annotation files as needed before evaluating {\tt \$URL}.) The program {\tt url\_view} is a shell script (normally installed at the same time as \WAVE{}, and in the same directory) that handles starting the browser if necessary and instructing it to display the specified URL. If {\tt URLV} is not set, or if you have set it to {\tt firefox}, {\tt url\_view} does so using the command: \begin{verbatim} ( firefox -remote 'openURL($URL)' || firefox $URL ) & \end{verbatim} \noindent In this case, \WAVE{} (via {\tt url\_view}) first uses Firefox's {\tt -remote} option to instruct an already-running copy of Firefox to open the desired URL. This is a very fast operation if Firefox is already running (since Firefox is already in RAM, it is not necessary to load another in order to run the {\tt firefox -remote ...} process, and the remote process efficiently delivers its message and exits immediately). If Firefox was not running, the {\tt firefox -remote} command fails, and (in this case only) {\tt url\_view} starts a new Firefox browser process. \index{Mozilla} The {\tt url\_view} script includes analogous commands for each of the other supported browsers. If {\tt url\_view} does not recognize the browser named by {\tt URLV}, it simply runs: \begin{verbatim} ( $URLV $URL ) & \end{verbatim} \noindent but this is (in general) much slower, since launching a new instance of the browser for each URL will take much more time than reusing an already-open window, as for the supported browsers. To configure \WAVE{} to use a different web browser, edit the {\tt url\_view} script appropriately. (You may also do this by editing the action associated with {\tt }, but it's better to modify {\tt url\_view}, since \WAVE{} also uses {\tt url\_view} to display some of its on-line help. If you decide to modify the action in the \WAVE{} menu file, be careful not to change the {\tt } tag at the beginning of the line, since \WAVE{} uses this tag to identify the browser interface command within the menu file.) Consult the documentation for your browser to see what commands will be needed. (The {\tt -remote} option is used by Firefox, Mozilla, Netscape, and Opera; Galeon and Konqueror use variations on this approach, and other browsers will probably require different methods.) Browsers that do not support any means of remote control are best avoided (since the only way to use them is to start a new browser for each new URL, a very inefficient solution). \index{Galeon}\index{Konqueror}\index{Mozilla}\index{Netscape}\index{Opera}\index{web browser} \chapter{Printing from \WAVE{}} \label{ch:printing} \section{Standard methods for printing} There are three standard ways to print annotated signals from \WAVE{}. The first is the easiest: \index{format!of printed output} \begin{itemize} \item From the {\sf File} menu, select {\sf Print}. This produces a ``chart recorder'' format copy of the contents of the signal window, as shown in figure~\ref{fig:chart1} \begin{latexonly} (page~\pageref{fig:chart1}). \end{latexonly} \end{itemize} The other two methods require that you first mark the region to be printed using the `{\tt <}' and `{\tt >}' markers, or by entering the time interval in the {\sf Start} and {\sf End} fields of the {\sf Analyze} window. You may also arrange the {\sf Signal list} so that it contains the signals you wish to print, in the desired top-to-bottom order. After this preparation, you are ready to print: \begin{itemize} \item Click on \button{Print chart} in the {\sf Analyze} window to produce a ``chart recording'' as shown in figure~\ref{fig:chart2}. \begin{htmlonly} \begin{rawhtml}

[Chart recording] \end{rawhtml} \end{htmlonly} \begin{figure} \begin{latexonly} \fbox{\centerline{\epsfig{file=chart2,width=.9\linewidth}}} \end{latexonly} \begin{htmlonly} ........................................................................_ \end{htmlonly} \caption{``Chart recording'' made using {\sf Print chart} from the {\sf Analyze} window.} \label{fig:chart2} \end{figure} \item Click on \button{Print full disclosure} in the {\sf Analyze} window to produce the highly compact format shown in figure~\ref{fig:full-disclosure}. \end{itemize} \begin{htmlonly} \begin{rawhtml}

[Full disclosure] \end{rawhtml} \end{htmlonly} \begin{figure} \begin{latexonly} \fbox{\centerline{\epsfig{file=fulldisc,width=.9\linewidth}}} \end{latexonly} \begin{htmlonly} ........................................................................_ \end{htmlonly} \caption{Plot made using {\sf Print full disclosure} from the {\sf Analyze} window.} \label{fig:full-disclosure} \end{figure} In addition to these three methods, you may also want to try the method described in \hyperref{Printing a Log} {section~} {, page~\pageref{sec:printing-log}, } {sec:printing-log} for printing a \WAVE{} log together with charts of the signals and annotations for each entry in it. \index{resolution!of printed output} In each case, printing is performed by spawning a {\tt pschart} or \index{analysis commands window@{\sf Analysis Commands} window} {\tt psfd} process in the {\sf Analysis Commands} window (after saving any edits). Since these commands reread the signals, the output generally has much higher resolution than is possible on screen. If you are unable to print from \WAVE{}, your printer may not be set up. See \hyperref{Printer Setup} {section~} {, page~\pageref{sec:printer-setup} } {sec:printer-setup} for instructions. \section{Creating custom formats for printing} \label{sec:custom-printing} \index{menu file} \index{pschart command@{\tt pschart} command} \index{psfd command@{\tt psfd} command} Since \WAVE{}'s printing facilities are implemented through its menu file, you have complete control over the format of the output, simply by editing the menu file as in the previous chapter. Both \htmladdnormallink{{\tt pschart}}{../wag/pschar-1.htm} (used to produce ``chart recorder'' output) and \htmladdnormallink{{\tt psfd}}{../wag/psfd-1.htm} (used to produce ``full disclosure'' output) are highly flexible and can be configured to produce almost any reasonable format by inserting the appropriate options in the commands given in \WAVE{}'s menu file. You may change the predefined commands or add new ones as you wish. \index{Print tag in menu file@{\tt } tag in menu file} In the menu file, the special tag ``{\tt }'' identifies the command that runs when you select {\sf Print} from the {\sf File} menu. You can change this command as you wish, but don't change the ``{\tt }'' tag itself. A few of the most commonly-used options (usable with either {\tt pschart} or {\tt psfd}) are listed below. For further details on these and other options, see \htmladdnormallink{{\tt pschart(1)}}{../wag/pschar-1.htm} and \htmladdnormallink{{\tt psfd(1)}}{../wag/psfd-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. (You can also get a brief summary of the many available options by typing {\tt pschart~-h} and {\tt psfd~-h}.) \begin{description} \index{printer!resolution} \index{resolution!of printed output} \item[{\tt -d} \textit{n}] Optimize for use with a printer with a resolution of \textit{n} dots per inch (default: 300 dpi, the typical resolution for low-end laser printers). This option has no effect on the output scale, and need not be correct for your printer; it does, however, affect the amount of detail rendered and the thickness of the lines that are drawn. \item[{\tt -E}] \index{PostScript} Generate EPSF format, suitable for inclusion in another PostScript file. (I used this option when preparing the {\tt pschart} and {\tt psfd} plots in this guide). The default format is PostScript, but \emph{not} EPSF, for two reasons. First, a legal EPSF plot must fit entirely on one page, but {\tt pschart} and {\tt psfd} are frequently used to produce multi-page plots. Second, an EPSF plot may be rescaled if it is included in another document, and the scales printed at the bottom of the plot will be incorrect if this happens (as is the case in the plots reproduced in this guide). \item[{\tt -g}] Print a grid (using {\tt pschart}) or a set of time axes (using {\tt psfd}). \item[{\tt -l}] Label the signals in the plot margins. \item[{\tt -L}] Print in ``landscape'' orientation (default: portrait orientation). The charts made by selecting {\sf Print} from the {\sf File} menu are printed using this option. \item[{\tt -M}\textit{n}] The numeric argument \textit{n} attached to this option (there should be no whitespace between {\tt -M} and \textit{n}) controls where annotations appear relative to the signals, and the appearance of the marker bars. Unless you display annotations as a signal in \WAVE{}, you may wish to use the option {\tt -M\$DISPMODE} in the menu file -- if you do so, then your printed output will show annotations and marker bars the same way that \WAVE{} does. \item[{\tt -P} \textit{pagesize}] Use this option to print on non-standard paper, or to make a plot that fits in a smaller space. The default \textit{pagesize} is set when {\tt psfd} and {\tt pschart} are compiled. For the precompiled binaries distributed with \WAVE{}, the default is {\tt letter} (US letter size paper, 8.5 x 11 inches, or 216 x 279 mm). In most of the rest of the world, the standard is {\tt A4} (210 x 297 mm). You can specify pagesize as ``{\tt letter}'', ``{\tt A4}'', or a number of other popular sizes, or as ``\textit{width}x\textit{height}'', where \textit{width} and \textit{height} are the dimensions in millimeters. I used this option to prepare figure~\ref{fig:chart1} \begin{latexonly} (on page~\pageref{fig:chart1}) \end{latexonly} in a relatively compact form; if you print a figure like it with the default page size, there will be considerable empty space between the plot itself and the page headers and footers. \item[{\tt -t} \textit{n}] This option sets the time scale, in mm/sec. Defaults are 12.5 mm/sec for {\tt pschart}, and 2.5 mm/sec for {\tt psfd}. For {\tt pschart}, you might wish to use the option {\tt -t \$TSCALE} in the menu file, so that your charts will be printed at the same time scales you choose for \WAVE{}'s signal window. \item[{\tt -v} \textit{n}] \index{amplitude scale!on printed output} This option sets the amplitude scale, in mm/mV. Defaults are 5 mm/mV for {\tt pschart}, and 1 mm/mV for {\tt psfd}. For {\tt pschart}, you might wish to use the option {\tt -v \$VSCALE} in the menu file, so that your charts will be printed at the same amplitude scales you choose for \WAVE{}'s signal window. \end{description} \index{PostScript} \index{printer!non-PostScript} This list of options only hints at what is possible using {\tt pschart} and {\tt psfd}. There are many other options described in the {\tt man} pages, and it is also possible to redefine the PostScript primitives used by these programs (for example, to change the appearance of the grid, or to plot in color on a color-capable printer) by modifying their PostScript prolog files. It is also possible to replace these programs entirely (for example, with your own applications for printing on a non-PostScript printer) if desired, simply by editing \WAVE{}'s menu file. \chapter{Calibration} \label{ch:calibration} \index{calibration}\index{display calibration}\index{signal!calibration} \begin{htmlonly} \index{adus (analog/digital converter units)} \end{htmlonly} \begin{latexonly} \index{adus (analog/digital converter units)|textbf} \end{latexonly} \index{physical units} \index{units!physical}\index{units!ADC} By \emph{calibration}, we refer to the information needed to establish the true amplitudes of signals. There are two distinct aspects of calibration. First is \emph{signal calibration}, the process of establishing the relationship between the sample values in analog-to-digital converter units (\emph{adus}) and the physical units of the signals (millivolts, millimeters of mercury, etc.). Second is \emph{display calibration}, the process of determining standard scales for displaying various types of signals (e.g., 10 mm/mV, 1 mm/10 mmHg). \section{Signal calibration} \index{header file}\index{gain}\index{signal!gain}\index{signal!baseline} \index{baselines} If your signals come from standard databases (see \hyperref{System Requirements} {Appendix~} {, page~\pageref{app:system-requirements}} {app:system-requirements}), signal calibration is usually not a concern. Briefly, the \emph{header file} for each database record contains calibration information for each signal. This information includes (for each signal) the gain (the number of adus per physical unit), the baseline (the sample value that corresponds to zero physical units), the type of physical unit, and the type of signal. (Not all header files contain all of this information, but \WAVE{} can make reasonable guesses about any of it that may be missing in most cases.) If you have digitized your own signals, however, you should generally calibrate them before processing them further. \WAVE{} makes it easy to do so, provided that you have recorded signals with known amplitudes. The procedure is: \begin{itemize} \item Use \WAVE{} to display the uncalibrated signals. \index{calibration pulse} \item Find a calibration pulse (or a segment with known amplitudes). Mark samples of the low and high amplitude phases using the `{\tt <}' and `{\tt >}' markers respectively. \begin{figure} \centerline{\epsfig{file=calibrate}} \caption[Signal calibration.]{Signal calibration using \WAVE{}. Signal 3 (ABP) has been selected (note that all of the other signal numbers have been removed from the {\sf Signal list}). The `{\tt <}' and `{\tt >}' markers bracket a step in the recorded calibration pulse near the center of the signal window. In the {\sf Analyze} window, we have clicked on \button{Calibrate}, and the result of doing so appears in the {\sf Analysis Commands} window.} \label{fig:calibration} \index{calibration} \begin{htmlonly} \index{analysis commands window@{\sf Analysis Commands} window} \end{htmlonly} \begin{latexonly} \index{analysis commands window@{\sf Analysis Commands} window|textit} \end{latexonly} \end{figure} \item Set the {\sf Signal list} field in the {\sf Analyze} window to match the signal(s) you have marked. You can calibrate more than one signal at a time if suitable calibration pulses for each signal are present between the `{\tt <}' and `{\tt >}' markers. Be sure to remove any other signals from the {\sf Signal list}. \item Click left on \button{Calibrate} in the {\sf Analyze} window, and answer any questions that appear in the {\sf Analysis Commands} window. These questions will always appear if the header file for the record does not contain signal descriptions or physical unit specifications; some or all of the questions may not appear if the signal type and calibration pulse type are already known. \item Repeat the previous steps for any other signals that require calibration. \item Close and reopen the record. (\WAVE{} does not adjust the display scales until this has been done.) \end{itemize} \index{signal!calibration}\index{calibration!signal} Figure~\ref{fig:calibration} illustrates signal calibration using a record from the MIMIC Database. In this case, the type of signal (ABP) is known from an entry in the {\tt WFDBCAL} file (not shown here), so that {\tt calsig} is able to determine the physical units of the signal (mmHg). Since a variety of calibration pulses are used for ABP signals, the {\tt WFDBCAL} file does not specify the pulse levels, which {\tt calsig} has asked us to enter (in this case, we have entered {\tt 60} and {\tt 120}). Based on this information, {\tt calsig} determines the offset and gain needed to convert raw sample values for signal 3 into ABP measurements in mmHg. {\tt calsig} then makes the appropriate changes to the header file for the current record. When \WAVE{} or another WFDB application next opens this record, the ABP signal will be properly calibrated. \index{calsig command@{\tt calsig} command} By default, {\tt calsig} generates an amplitude histogram of the samples between the `{\tt <}' and `{\tt >}' markers. It then identifies the low and high amplitude portions of the calibration pulse by searching for the two largest distinct modes in this amplitude histogram. For this reason, {\tt calsig} works best if the segment bounded by the `{\tt <}' and `{\tt >}' markers includes at least a few samples of both the high and low amplitude phases. Avoid placing either marker immediately next to the transition point between the phases if possible. If {\tt calsig} fails to find two distinct peaks in the amplitude histogram, it will produce an error message; if this happens, adjust the positions of the markers and try again. In some cases (for example, if the calibration pulse is a sawtooth, as in the RESP signal at the bottom of figure~\ref{fig:calibration}), this strategy may fail, no matter where the markers are placed. In such cases, try again with {\tt calsig}'s {\tt -q} or {\tt -Q} options to use one of its alternate algorithms. (These are less robust since they depend on differential rather than integrative measurements, but they can be used in a pinch.) For further information on signal calibration, see \htmladdnormallink{{\tt calsig(1)}}{../wag/calsig-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. \section{Display calibration} \index{display calibration}\index{calibration!display} Display calibration is performed by \WAVE{} in several steps: \begin{itemize} \index{resolution!of display} \index{dpi option for WAVE@{\tt -dpi} option for \WAVE{}} \index{command-line options!{\tt -dpi}}\index{options!{\tt -dpi}} \item First, \WAVE{} determines the display resolution (either from the {\tt -dpi} com\-mand-line option if possible, otherwise from the {\tt Wave.Dpi} resource if possible, or by querying the X server if neither the command-line option nor the resource was specified). \index{scales!time}\index{scales!amplitude}\index{time scale} \index{amplitude scale}\index{View window@{\sf View} window} \item Next, \WAVE{} determines what scales to use for time and amplitude. You can set these using the \amenubutton{Time scale:} and \amenubutton{Amplitude scale:} menu buttons in the {\sf View} window. The amplitude scale refers to standard ECG signals; the default scales (10 mm/mV and 25 mm/sec) are those used for standard ECG chart recordings. If you choose a non-standard amplitude scale, \WAVE{} sets its magnification factor to the ratio between that scale and 10 mm/mV. \item Finally, if any of the signals to be displayed are not ECGs, \WAVE{} checks the WFDB calibration file (named by the {\tt WFDBCAL} environment variable) for entries that specify the standard display scales for those signals. \WAVE{} multiplies these scales by its magnification factor (see above) in order to determine the final display scales. \end{itemize} \index{signal!type}\index{calibration file} \index{WFDBCAL environment variable}\index{environment variables!WFDBCAL} \index{Load window@{\sf Load} window} In general, it is not necessary to do anything to make \WAVE{} display common signals at reasonable scales. If you have a previously undefined signal type, or you need to change the size of a particular type of signal relative to that of an ECG, however, simply edit a copy of the WFDB calibration file, reset {\tt WFDBCAL}, and restart \WAVE{}. (If you wish, you need not exit and restart \WAVE{}; instead, give the copy a different name, then change the {\sf Calibration file} item in the {\sf Load} window to match. Any changes are effective only when the signal window is redrawn.) See \htmladdnormallink{{\tt wfdbcal(5)}}{../wag/wfdbca-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details. If the display calibration appears incorrect for all signals, the X server may be supplying incorrect information about the display size to \WAVE{}. This problem can be diagnosed by setting the time and amplitude scales to their default values (nominally 25 mm/sec and 10 mm/mV), and displaying the default grid (0.5 mV x 0.2 s). If the grid intervals are not 5 mm in each direction, follow the procedure in \hyperref{``How can I get correct display scales?''} {section~} {, page~\pageref{faq:display-scales}} {faq:display-scales}. \chapter{Customizing \WAVE{}} In the previous chapters, we saw how \WAVE{}'s {\sf Analyze} window can be modified by editing the \WAVE{} menu file. This chapter describes several other ways in which \WAVE{} can be customized. \section{Using the {\sf View} window} \begin{latexonly} In section~\ref{sec:selecting-annotation}, we briefly opened the {\sf View} window (see figure~\ref{fig:view-window}, page~\pageref{fig:view-window}) in order to turn on the marker bar display. Open the {\sf View} window again, and examine the other controls in it. \end{latexonly} \begin{htmlonly} Earlier, we briefly opened the {\sf View} window (see figure~\ref{fig:view-window}) in order to turn on the marker bar display. Open the {\sf View} window again, and examine the other controls in it. \end{htmlonly} Three important controls are located along the bottom edge of the window. \button{Redraw} makes \WAVE{} refresh the signal window, and also closes the {\sf View} window You should click on \button{Redraw} in order to see the effects of any changes you make in the {\sf View} window. The other two buttons in this group are described below. Along the top of the window, just below the title bar, is a row of on/off controls used to turn on various optional display elements. These controls appear depressed (darker than the background) when they are `on'. The first four of these ({\sf subtype}, {\sf `chan' field}, {\sf `num' field}, and {\sf `aux' field}) control the display of secondary fields in the annotations. If any of these controls are `on', the corresponding fields are displayed below each annotation mnemonic, in the order shown in the {\sf View} window. \index{amplitude scale} \index{time scale} The \amenubutton{Time scale:} and \amenubutton{Amplitude scale:} menu buttons allow you to choose any of a wide range of standard scales for the signal (and {\sf Scope}) windows. Click right on the \button{$\nabla$} to open the menu. The \amenubutton{Draw:} menu offers the choice of displaying {\sf all signals} (the default) or {\sf listed signals only} (i.e., those named in the {\sf Signal list} in the {\sf Analyze} window). By choosing to display {\sf listed signals only}, you may rearrange the signals within in the signal window. By listing a signal in two or more entries in the signal list, you can arrange to have that signal drawn in two or more locations; this can be useful for making side-by-side comparisons of a signal against several others. You may also find it useful to remove one or more signals from the signal list in order to reduce crowding in the signal window; this technique is nearly essential if the record has more than about 12 signals. (You can also change the spacing between signals uniformly by resizing \WAVE{}'s main window.) \index{format!of annotation display} The \amenubutton{Show annotations:} menu offers three ways to display annotations: {\sf centered} (the default), {\sf attached to signals}, and {\sf as a signal}. If you choose {\sf attached to signals}, the {\tt chan} field of each annotation specifies the signal number to which the annotation is attached, and the annotation is displayed slightly above this signal. (Any annotations that have {\tt chan} fields that are not valid signal numbers for the current record are displayed at the center of the signal window.) If you choose {\sf as a signal}, the {\tt num} field of each annotation is taken as the amplitude of a signal at the time of the annotation, and \WAVE{} draws this signal in the center of the signal window by connecting the amplitude/time pairs specified by the annotations with straight line segments. \index{format!of time display/entry} \index{absolute time} \index{elapsed time} \index{time!display mode} The \amenubutton{Time display:} menu provides three alternatives for display of times in the lower corners of the signal window: {\sf elapsed} (the default), {\sf absolute}, and {\sf in sample intervals}. Elapsed time measures the interval in hours, minutes, and seconds from the beginning of the record. Absolute time (i.e., time of day and date) is not defined for all records. If it is available, choosing {\sf absolute} from this menu will cause \WAVE{} to show absolute times in the signal window, using the color \WAVE{} uses for annotations; otherwise, \WAVE{} switches to {\sf elapsed} time display. Times may also be displayed {\sf in sample intervals} (counting from the beginning of the record) for any record; these may be recognized in the signal window by the `{\sf s}' prefix. \index{format!of grid display} The choice you make from the \amenubutton {Grid display:} menu determines how \WAVE{} draws the background grid in the signal window. The grid intervals are fixed in time and amplitude units. At the default scales (25 mm/sec, 10 mm/mV) the default grid ({\sf 0.2 s x 0.5 mV}) has 5 mm intervals if the display is properly calibrated. (If this is not the case, see \hyperref{``How can I get correct display scales?''} {section~} {, page~\pageref{faq:display-scales}} {faq:display-scales}.) If you change the time or amplitude scales, the grid intervals change size to match, so that you always have a visual cue about the display scale if you leave the grid display on. If you choose a very small time scale (i.e., one that permits \WAVE{} to display a large amount of data in the signal window), the grid may appear solid grey; in this case, you may wish to choose ``{\sf 1 m x 0.5 mV}'' from this menu (so that the vertical grid lines appear at 1-minute intervals), or `{\sf 0.5 mV}'' (thereby suppressing the vertical grid lines), or even ``{\sf no grid}''. At relatively large scales, where it may be useful to have a finer grid, choose ``{\sf 0.04 s x 0.1 mV}''. To display highly time-compressed data with fine horizontal grid lines, choose ``{\sf 1 m x 0.1 mV}''. If you make changes in the {\sf View} window and wish to discard them before clicking on \button{Redraw}, you may do so by clicking on \button{Undo changes}. This button does not restore the initial values if you have registered earlier changes by using \button{Redraw}; to do this, you must restart \WAVE{}. If, on the other hand, you have made changes in the {\sf View} window and wish to have \WAVE{} start up with these settings, click on \button{Save as new defaults}. Note that \WAVE{} saves the current state of the signal window \emph{as it appears when you click on} \button{Save as new defaults}; if you haven't made your changes effective using \button{Redraw}, they won't be saved. \subsection{Highly time-compressed displays} At highly compressed time scales (defined as any scale of 125 mm per minute or less), \WAVE{} normally suppresses the grid display, even if you have specified a default grid display for use at other time scales. To change this behavior, select the desired grid display mode, click on \button{Redraw}, and then on \button{Save as new defaults} \emph{while viewing signals at 125 mm/min or less}. \subsection{Low-rate records} Low-rate records (defined as those that are sampled at rates of 10 Hz or less) are displayed by default at a time scale of 5 mm per minute. To change this default, simply change to the desired scale (all of the standard scales are available in any case) click on \button{Redraw}, and then on \button{Save as new defaults} \emph{while a low-rate record is open}. \section{Resizing the signal and {\sf Scope} windows} \WAVE{}'s main window, which contains the signal window, may be resized using the resize ``handles'' provided by your window manager. If you are using an Open Look window manager such as {\tt olwm} or {\tt olvwm}, as recommended, the resize handles are at the corners of the window. To use them, move the pointer over a resize handle; when the pointer changes shape, click left and drag it to the desired location. (\WAVE{} adjusts the window width to the nearest multiple of one second.) If you use \button{Save as new defaults} in the {\sf View} window, as described above, the new window size will be the default for future sessions. The {\sf Scope} window may be resized in the same way, but its default (initial) size may not be changed. \section{\WAVE{}~environment variables} \index{environment variables!list of} \WAVE{} uses many environment variables; they are listed in this section roughly in order of importance. Many of them need not be set at all, since \WAVE{} uses reasonable default values in most cases. Those that are set must be set on the \WAVE{} host. \index{WAVE host@\WAVE{} host} For general information on how to set environment variables, see the discussion of the \hyperref{{\tt DISPLAY} variable} {{\tt DISPLAY} variable in section~} {, page~\pageref{sec:setting-display}} {sec:setting-display}. \begin{description} \index{DISPLAY environment variable}\index{environment variables!DISPLAY} \item[{\tt DISPLAY}] The name of the X server and display you are using (see \hyperref{the worksheet} {section~} {, page~\pageref{worksheet}} {worksheet}). If not set, \WAVE{} will not be able to open windows on your display unless your display is attached to the \WAVE{} host directly (not via network connection). \index{WFDB environment variable}\index{environment variables!WFDB} \item[{\tt WFDB}] The database path (see \htmladdnormallink{{\tt setwfdb(1)}}{../wag/setwfd-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. If not set, \WAVE{} can find database files only in the locations specified by the compiled-in default database path, which normally includes the current directory and the {\tt /usr/database} directory (on the \WAVE{} host), and PhysioBank's archives at \begin{htmlonly} \htmladdnormallink{\tt http://physio\-net.\-org/\-physio\-bank/\-data\-base}{http://physionet.org/physiobank/database/} \end{htmlonly} \begin{latexonly} {\tt http://physio\-net.\-org/\-physio\-bank/\-data\-base} \end{latexonly} (if the Internet is accessible from the \WAVE{} host). \index{WFDBCAL environment variable}\index{environment variables!WFDBCAL} \item[{\tt WFDBCAL}] The WFDB calibration file (see \htmladdnormallink{{\tt setwfdb(1)}}{../wag/setwfd-1.htm} and \htmladdnormallink{{\tt wfdbcal(5)}}{../wag/wfdbca-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, and \hyperref{Calibration} {chapter~} {, page~\pageref{ch:calibration}} {ch:calibration}). If not set, \WAVE{} reads a default calibration file specified when the WFDB library was compiled. If the calibration file cannot be read, \WAVE{} may not scale signals other than ECGs correctly. \index{WAVEMENU environment variable}\index{environment variables!WAVEMENU} \item[{\tt WAVEMENU}] The name of \WAVE{}'s \hyperref{menu file} {menu file (see section~} {, page~\pageref{sec:menu-file}} {sec:menu-file}). If not set, \WAVE{} uses {\tt wavemenu} in the current directory if possible; otherwise, it uses \WAVE{}'s default menu ({\tt \$MENUDIR/wavemenu.def}). \index{PATH environment variable}\index{environment variables!PATH} \index{SHELL environment variable}\index{environment variables!SHELL} \index{analysis commands window@{\sf Analysis Commands} window} \item[{\tt SHELL}] The command interpreter used within the {\sf Analysis Commands} window. If not set, \WAVE{} uses {\tt /bin/sh} (the Bourne shell). Other shell-related variables, such as {\tt PATH}, are also significant when \WAVE{} is running commands within the {\sf Analysis Commands} window. \index{EDITOR environment variable}\index{environment variables!EDITOR} \item[{\tt EDITOR}] The name of the text editor to be used for modifying \WAVE{}'s menu file. If not set, \WAVE{} uses {\tt textedit}. \index{URLV environment variable}\index{environment variables!URLV} \item[{\tt URLV}] The name of the web browser to be used for viewing LINK annotation attachments and some of \WAVE{}'s on-line help; normally, {\tt URLV} should be one of {\tt galeon}, {\tt konqueror}, {\tt mozilla}, {\tt netscape}, or {\tt opera}. If not set, \WAVE{} uses Mozilla. See \hyperref{\WAVE{} and the Web} {section~} {, page~\pageref{sec:web},} {sec:web} for details. \index{Galeon}\index{Konqueror}\index{Mozilla}\index{Netscape}\index{Opera} \index{web browser} \index{PRINTER environment variable}\index{environment variables!PRINTER} \index{PostScript} \item[{\tt PRINTER}] The name of a PostScript printer accessible to the \WAVE{} host, to be used for paper output. If not set, \WAVE{} uses the default printer. \index{PSPRINT environment variable}\index{environment variables!PSPRINT} \item[{\tt PSPRINT}] \index{lpr command@{\tt lpr} command} The command needed to print PostScript from the standard input. If not set, \WAVE{} uses the command {\tt lpr} (or {\tt lpr -P\$PRINTER}, if {\tt PRINTER} is set). \index{TEXTPRINT environment variable}\index{environment variables!TEXTPRINT} \item[{\tt TEXTPRINT}] The command needed to print ordinary text from the standard input. If not set, \WAVE{} uses the command {\tt lpr} (or {\tt lpr -P\$PRINTER}, if {\tt PRINTER} is set). \index{ANNTAB environment variable}\index{environment variables!ANNTAB} \index{annotation!table of definitions} \item[{\tt ANNTAB}] The name of a file that contains custom annotation definitions (see the discussion of \hyperref{{\tt Wave.Anntab}} {{\tt Wave.Anntab} in section~} {, page~\pageref{anntab}} {anntab} for details). If not set, \WAVE{} uses standard annotation definitions only. \end{description} The environment variables below are not needed unless the \WAVE{} binary distribution, or XView, \index{XView} has been installed in non-standard directories: \begin{description} \index{on-line help} \index{HELPPATH environment variable}\index{environment variables!HELPPATH} \item[{\tt HELPPATH}] The path for XView spot help; if not set, \WAVE{} initializes it to {\tt /usr/lib/help}. \WAVE{}'s own spot help is in {\tt \$HELPDIR/wave}, which is appended to the end of {\tt HELPPATH} by \WAVE{}. \index{HELPDIR environment variable}\index{environment variables!HELPDIR} \item[{\tt HELPDIR}] The directory in which \WAVE{}'s help directory is located; if not set, \WAVE{} uses {\tt /usr/help}. \index{MENUDIR environment variable}\index{environment variables!MENUDIR} \item[{\tt MENUDIR}] The name of the directory that contains \WAVE{}'s default menu file; if not set, \WAVE{} uses {\tt /usr/lib}. \index{RESDIR environment variable}\index{environment variables!RESDIR} \item[{\tt RESDIR}] The name of the directory in which system-wide default X11 resource files are kept; if not set, \WAVE{} uses {\tt /usr/lib/X11/app-defaults}. See the {\tt man} page for \WAVE{} for information about other directories that are searched for resource files. \end{description} \section{X11 resources for \WAVE{}} \label{sec:resources} \index{resources}\index{X11 resources} You can control many aspects of \WAVE{}'s appearance and behavior by setting its resources. If you are not familiar with this concept, refer to an introductory book on using the X Window System, such as Quercia and O'Reilly's {\it X Window System User's Guide}. Since \WAVE{} is built using the XView toolkit, \index{XView} all of the resources listed in the {\tt man} page for {\tt xview} can be used with \WAVE{} (type `{\tt man xview}' for details). If your system doesn't have an XView {\tt man} page, refer to the copy provided with the \WAVE{} distribution (\htmladdnormallink{{\tt xview.7}}{../wag/xview-7.htm}). In addition, the \WAVE{}-specific resources listed below may also be set. The standard way to change default values of X11 resources is to define them in a file named {\tt .Xdefaults} (note the initial `{\tt .}') in your home directory. \WAVE{} does this when you use \button{Save as new defaults} in the {\sf View} panel. If you use \WAVE{} on workstations with different display capabilities, you can create custom resource settings for each one by moving these resource definitions from {\tt .Xdefaults} to {\tt .Xdefaults-{\it hostname}} (where {\it hostname} specifies the system to which the display is attached). \begin{description} \index{annotation!marker bars}\index{marker bars} \item[{\tt Wave.AllowDottedLines}] This resource specifies if \WAVE{} is allowed to render dotted lines. \WAVE{} normally draws annotation marker bars as dotted lines, and may use dotted lines for other display elements on black-and-white displays for clarity. Some X servers do not properly render dotted lines, however; if you observe irregular or missing annotation marker bars, change the value of this resource from `{\tt True}' to `{\tt False}' (by editing {\tt .Xdefaults}). \label{anntab} \index{ANNTAB environment variable}\index{environment variables!ANNTAB} \index{annotation!table of definitions} \item[{\tt Wave.Anntab}] This resource specifies the name of a file that contains a table of annotation definitions. The environment variable {\tt ANNTAB} can also be used to specify this filename; the resource overrides the environment variable if both are set. The file contains one-line entries of the form \begin{verbatim} 15 % Funny looking beat \end{verbatim} in which the first field specifies the (numeric) annotation code in the range between 1 and {\tt ACMAX} inclusive (see {\tt /usr/include/wfdb/ecgcodes.h} for a list of predefined codes and for the definition of {\tt ACMAX}); the second field (`{\tt \%}' in the example) is a mnemonic (used in annotation display and entry), and the remainder of the entry is a description of the intended use of the annotation code (which appears next to the mnemonic in the {\sf Type} field and menu of the {\sf Annotation Template} window). Lines in the annotation table that begin with `{\tt \#}' are treated as comments and ignored. It is not necessary to specify an annotation table when editing an existing annotation file unless previously undefined annotation types are to be added to it during the editing process, although it is generally harmless to do so. \index{resolution!of display} \index{display resolution}\index{resolution} \index{dpi option for WAVE@{\tt -dpi} option for \WAVE{}} \index{command-line options!{\tt -dpi}}\index{options!{\tt -dpi}} \item[{\tt Wave.Dpi}] This resource specifies the display resolution in dots per inch in the form {\it mm}{\tt x}{\it nn}, where {\it mm} is the horizontal resolution and {\it nn} is the vertical resolution. Normally, the resolution is known to the X server, and it is unnecessary to specify this resource. If your X server is misinformed, \WAVE{}'s calibrated display scales will be incorrect; the best solution is to specify the resolution using a server option such as the {\tt -dpi} option supported by MIT's X11 servers, since this will solve problems common to any other applications that require calibrated scales as well. Not all X11 servers support such an option, however, so this option is available as a work-around. The command-line option {\tt -dpi} overrides the resource if both are specified. \index{graphics mode} \index{display mode} \index{g option for WAVE@{\tt -g} option for \WAVE{}} \index{command-line options!{\tt -g}}\index{options!{\tt -g}} \index{m option for WAVE@{\tt -m} option for \WAVE{}} \index{command-line options!{\tt -m}}\index{options!{\tt -m}} \index{O option for WAVE@{\tt -O} option for \WAVE{}} \index{command-line options!{\tt -O}}\index{options!{\tt -O}} \index{S option for WAVE@{\tt -S} option for \WAVE{}} \index{command-line options!{\tt -S}}\index{options!{\tt -S}} \item[{\tt Wave.GraphicsMode}] This resource specifies the graphics mode used by \WAVE{}; it can be overridden using the {\tt -g}, {\tt -m}, {\tt -O}, or {\tt -S} options. The legal values are 1 (monochrome mode), 2 (overlay greyscale mode), 4 (shared color mode), 6 (shared grey mode), and 8 (overlay color mode). \label{setting-colors} \index{colors!signal window} \item[{\tt Wave.SignalWindow.\{Grey|Color\}.}{\it element}] These resources specify the colors to be used on greyscale or color displays. The {\tt Color.}{\it *} resources are used only if the display is color-capable and neither greyscale nor monochrome mode has been specified. {\it element} is one of {\tt Background}, {\tt Grid}, {\tt Cursor}, {\tt Annotation}, or {\tt Signal}. The defaults are: \begin{center} \begin{tabular}{lll} {} & {\tt Grey} & {\tt Color} \\ \hline {\tt Background} & white & white \\ {\tt Grid} & grey75 & grey90 \\ {\tt Cursor} & grey50 & orange red \\ {\tt Annotation} & grey25 & yellow green \\ {\tt Signal} & black & blue \\ \end{tabular} \end{center} \item[{\tt Wave.SignalWindow.Mono.Background}] In monochrome mode, the background is normally white, and all other elements are normally black. The reverse can be obtained by setting this resource to `{\tt black}'. (There is at least one X server for which this fails.) \index{Scope window@{\sf Scope} window} \index{colors!scope window} \item[{\tt Wave.Scope.\{Grey|Color\}.\{Foreground|Background\}}] These resources specify the colors to be used in the {\sf Scope} window on greyscale or color displays. The {\tt Foreground} color is used for the waveform and the time display; by default, it matches the color used for signals in the signal window (see the previous item). Some X servers do not allow the background color of the {\sf Scope} window to be set, because of the color map animation and stippled erasing techniques used. \item[{\tt Wave.Scope.Mono.Background}] This resource can be used to invert the foreground and background of the {\sf Scope} window when \WAVE{} is running in monochrome mode. This does not work for all X servers. \index{signal!window!dimensions}\index{window!size} \item[{\tt Wave.SignalWindow.\{Height\_mm|Width\_mm\}}] These resources specify the preferred dimensions (in millimeters) for the signal window. The defaults are 120 and 250 respectively. \index{font!in signal window}\index{annotation!display font} \item[{\tt Wave.SignalWindow.Font}] This resource specifies the font used to display annotations and time marks in the signal window. The default is `{\tt fixed}'. \index{EDITOR environment variable}\index{environment variables!EDITOR} \item[{\tt Wave.TextEditor}] This resource specifies the name of the text editor invoked by \WAVE{} to permit you to edit \WAVE{}'s log and analysis menu files. The default is {\tt textedit} (the OpenLook visual editor). You may override this resource by using the environment variable {\tt EDITOR}, which is also used by many other UNIX applications that invoke editors. \item[{\tt Wave.View.{\it showoption}}] These resources specify which of the {\sf Show:} options at the top of the {\sf View} window are enabled by default. {\it showoption} is one of {\tt Subtype}, {\tt Chan}, {\tt Num}, {\tt Aux}, {\tt Markers}, {\tt SignalNames}, {\tt Baselines}, or {\tt Level}. The values of these resources are `{\tt True}' or `{\tt False}'. \index{amplitude scale} \index{time scale} \item[{\tt Wave.View.{\it menuname}}] These resources specify the positions of the initial choices in the corresponding {\sf View} menus, where the top item on each menu is in position 0, the one below it is in position 1, etc. {\it menuname} is one of {\tt TimeScale}, {\tt AmplitudeScale}, {\tt SignalMode}, {\tt AnnotationMode}, {\tt AnnotationOverlap}, {\tt TimeMode}, or {\tt GridMode}. For example, to set the initial time scale to 50 mm/sec (the item at position 13 in the \amenubutton{Time Scale:} menu), set {\tt Wave.View.TimeScale} to 13. \item[{\tt Wave.View.CoarseGridMode}] This resource specifies the initial grid display mode to be used at highly compressed time scales (125 mm/minute or less). The possible values are the same as those for {\tt Wave.View.GridMode}. The default is 0 (no grid); recommended alternatives are 5 (1 m x 0.5 mV) or 6 (1 m x 0.1 mV). \item[{\tt Wave.View.CoarseTimeScale}] This resource specifies the initial time scale to be used for low-rate records (those sampled at 10 Hz per signal or less). The possible values are the same as those for {\tt Wave.View.TimeScale}. The default is 5 (corresponding to 5 mm per minute). \end{description} \index{XView} \index{xrm option for WAVE@{\tt -xrm} option for \WAVE{}} \index{command-line options!{\tt -xrm}}\index{options!{\tt -xrm}} \index{default option for WAVE@{\tt -default} option for \WAVE{}} \index{command-line options!{\tt -default}}\index{options!{\tt -default}} In addition to the usual ways of setting X resources, it is possible to set any of those listed above, as well as any of the generic XView resources, by using the {\tt -xrm} or {\tt -default} options on the command line when starting \WAVE{}. For example, you can set the background color of the signal window using a command such as \begin{verbatim} wave -r 100s -xrm Wave.SignalWindow.Color.Background:lightblue \end{verbatim} \appendix \chapter{Summary of \WAVE{}~controls} This appendix contains an entry for each of \WAVE{}'s controls. The entries appear according to the position of the controls in left-to-right order within each window or menu. The {\sf Annotation Template}, {\sf Search Template}, and {\sf Scope} window controls are listed separately at the end. \index{help}\index{on-line help} This information is also available as spot help for each control (point to the control and press the \keycap{HELP} key, or the \keycap{F1} key if your keyboard does not have a \keycap{HELP} key). \section{\WAVE{}'s main window} \begin{figure}[h] \centerline{\epsfig{file=main-control-panel}} \end{figure} \begin{description} \index{File menu@{\sf File} menu} \item[\menubutton{File}] This button opens the {\sf File} menu, containing selections for loading, saving, printing, analyzing, and logging database files. (See \hyperref{File menu} {section~} {, page~\pageref{sec:file-menu}} {sec:file-menu}.) \index{View window@{\sf View} window} \item[\button{View...}] This button pops up the {\sf View} window, which allows you to choose (or merely examine) display scales, grid styles, and annotation, signal, and time display styles. Changes are not effective until the signal window is redrawn. (See \hyperref{View window} {section~} {, page~\pageref{sec:view-window}} {sec:view-window}.) \index{Edit menu@{\sf Edit} menu} \item[\menubutton{Edit}] The {\sf Edit} menu allows you to specify if annotation editing is to be allowed or forbidden. By default, editing is forbidden when \WAVE{} starts up. (See \hyperref{Edit menu} {section~} {, page~\pageref{sec:edit-menu}} {sec:edit-menu}.) \index{Properties menu@{\sf Properties} menu} \item[\menubutton{Properties}] This button brings up the {\sf Properties} menu, with selections for obtaining information about the current signal and annotation files and about this version of \WAVE{}. (See \hyperref{Properties menu} {section~} {, page~\pageref{sec:prop-menu}} {sec:prop-menu}.) \index{Find window@{\sf Find} window} \index{search}\index{annotation!searching for} \item[\button{{\tt <} Search}] This button recenters the signal window on the previous occurrence of an annotation or marker that matches the entry in the {\sf Search for} field of the {\sf Find} window, if any. If no match is found, a notice is posted, but the signal window is not recentered. If the {\sf Search for} field is empty, any annotation or marker will be counted as a match. \item[\button{\tt <<}] This button scrolls the signal window towards the beginning of the record, by an amount equal to the width of the signal window (i.e., a full screen). \item[\button{\tt <}] This button scrolls the signal window towards the beginning of the record, by an amount equal to half of the width of the signal window. \index{Find window@{\sf Find} window} \index{search}\index{annotation!searching for} \item[\button{Find...}] This button opens the \hyperref{{\sf Find} window} {{\sf Find} window (see section~} {, page~\pageref{sec:find-window})} {sec:find-window}, which allows you to specify what portion of the current record should be displayed next. You may set a specific start or end time, or you may specify an annotation to be searched for by \button{{\tt <} Search} and \button{Search {\tt >}} buttons. \item[\button{\tt >}] This button scrolls the signal window towards the end of the record, by an amount equal to half of the width of the signal window. \item[\button{\tt >>}] This button scrolls the signal window towards the end of the record, by an amount equal to the width of the signal window (i.e., a full screen). \index{search}\index{annotation!searching for} \item[\button{Search {\tt >}}] This button recenters the signal window on the next occurrence of an annotation or marker that matches the entry in the {\sf Search for} field of the {\sf Find} window, if any. If no match is found, a notice is posted, but the signal window is not recentered. If the {\sf Search for} field is empty, any annotation or marker will be counted as a match. \index{on-line help}\index{Help Topics window@{\sf Help Topics} window} \item[\button{Help}] This button pops up the \hyperref{{\sf Help Topics} window} {{\sf Help Topics} window (see section~} {, page~\pageref{sec:help-topics-window})} {sec:help-topics-window}, containing buttons that name several subjects for which extensive on-line help is available. Choosing a topic allows you to browse through or print the associated help file in a scrollable text window. \index{annotator name!default} \index{Quit button@{\sf Quit} button} \item[\button{Quit}] \WAVE{} exits when you press this button, after saving your edits if any. If your input file would be overwritten as a result of saving your edits, it is first renamed by prefixing its annotator name with an underscore (`{\tt \_}'). If you have entered annotations but have not specified an annotator, the name by which you invoked this program (normally, `{\tt wave}') is used for the annotator name. \end{description} \section{The {\sf File} menu} \index{File menu@{\sf File} menu} \label{sec:file-menu} Open this menu using \menubutton{File} in \WAVE{}'s main window. \begin{description} \index{Load window@{\sf Load} window} \index{record!name}\index{annotator name}\index{calibration file} \index{database path} \item[{\sf Load}] This selection pops up the \hyperref{{\sf Load} window} {{\sf Load} window (see section~} {, page~\pageref{sec:load-window})} {sec:load-window}, in which you can enter a new record or annotator name, or change the name of the calibration file or the value of the database path within this window. \index{title bar!unsaved edit indicator} \index{annotation file!backup} \item[{\sf Save}] If there are unsaved edits, this selection saves them. The annotator name in the title bar is marked with parentheses if there are unsaved edits. Only one level of backup is preserved, so you will overwrite the original annotation file if it is in the current directory and you open the same annotator more than once. \index{printing!signal window contents} \index{screen dump} \item[{\sf Print}] This selection prints the contents of the signal window on paper. The output is made from the original signal files, and therefore is of better quality than a screen dump would be. Your edits, if any, are saved before printing, so that the output reflects any changes you have made. \index{printing!commands} \item[{\sf Print setup}] This selection pops up the \hyperref{{\sf Print setup} window} {{\sf Print setup} window (see section~} {, page~\pageref{sec:print-setup-window})} {sec:print-setup-window}, showing the commands \WAVE{} uses to print PostScript and text data from the standard input. \index{menu file} \index{Analyze window@{\sf Analyze} window} \index{WAVEMENU environment variable}\index{environment variables!WAVEMENU} \index{action buttons} \item[{\sf Analyze}] This selection pops up the \hyperref{{\sf Analyze} window} {{\sf Analyze} window (see section~} {, page~\pageref{sec:analyze-window})} {sec:analyze-window} containing a customizable set of action buttons, and the {\sf Analysis Commands} window (a {\tt cmdtool}-style terminal emulator). The names of the action buttons and their assigned actions are read from \WAVE{}'s menu file (by default, this is {\tt wavemenu}, if it exists in the current directory, otherwise {\tt /usr/lib/wavemenu.def}; the default may be overridden by setting the environment variable {\tt WAVEMENU} \index{environment variables!WAVEMENU} \index{WAVEMENU environment variable} to the name of a different file). The buttons are usually configured to perform various analysis functions on the current record; read the default menu file for details. \index{log file} \index{pschart command@{\tt pschart} command} \item[{\sf Log}] This selection pops up the \hyperref{{\sf Log} window} {{\sf Log} window (see section~} {, page~\pageref{sec:log-window})} {sec:log-window}, which allows you to name a log file, and to record in that file the current record name, the time of the samples at the center of the signal window, and (optionally) a one-line comment. Log files may be used as scripts for \htmladdnormallink{{\tt pschart}}{../wag/pschar-1.htm}. You may write to as many log files in a single session as you choose, and you may accumulate entries from multiple sessions in a single log file. \end{description} \section{The {\sf Load} window} \index{Load window@{\sf Load} window} \begin{figure}[h] \centerline{\epsfig{file=load-window}} \end{figure} \label{sec:load-window} Open this window by selecting {\sf Load} from the \menubutton{File} menu. \index{record!name} \begin{description} \item[{\sf Record}] Specifies the name of the record to be viewed. The initial value of this field is the name of the record that you specified on the command line. To view another record, select this field and enter another record name. \index{annotator name} \item[{\sf Annotator}] Specifies the name of the annotator whose annotations are shown. If you specified an annotator on the command line, the annotator name is the initial value of this field; otherwise, the field is initially empty. Fill in or change this field to view or edit a different set of annotations. \index{record!reloading} \item[\button{Reload}] Press this button to reload the current record. This is intended to be used if an external process has modified the record (e.g., by writing annotations, or by recalibrating a signal) since it was loaded into \WAVE{}. \index{calibration file} \item[{\sf Calibration file}] Specifies the name of the WFDB calibration file (a text file containing information on the relative scales of many types of signals). Initially, this field contains the value of the {\tt WFDBCAL} environment variable. \index{environment variables!WFDBCAL}\index{WFDBCAL environment variable} You should include path information in this field only if the WFDB calibration file is not found in the WFDB path. \index{database path} \index{WFDB environment variable}\index{environment variables!WFDB} \item[{\sf WFDB path}] Specifies the search path for \WAVE{}'s input files. Initially, this field contains the value of the {\tt WFDB} environment variable. \index{environment variables!WFDB}\index{WFDB environment variable} Components are directory names, separated by colons (`:'). An empty component (either an initial or final colon, or two consecutive colons) specifies the current directory. \end{description} \section{The {\sf Print setup} window} \index{Print setup window@{\sf Print setup} window} \begin{figure}[h] \centerline{\epsfig{file=print-setup-window}} \end{figure} \label{sec:print-setup-window} Open this window by selecting {\sf Print setup} from the \menubutton{File} menu. \begin{description} \item[{\sf PostScript print command}] \WAVE{} supplies any PostScript data to be printed (such as the chart generated when you choose {\sf Print} from the \menubutton{File} menu) to the standard input of this command. \item[{\sf Text print command}] \WAVE{} supplies any text to be printed (such as on-line help) to the standard input of this command. \end{description} \index{PRINTER environment variable}\index{environment variables!PRINTER} \index{PSPRINT environment variable}\index{environment variables!PSPRINT} \index{TEXTPRINT environment variable}\index{environment variables!TEXTPRINT} The initial settings in the {\sf Print setup} window are determined by the environment variables {\tt PSPRINT} and {\tt TEXTPRINT}; if either of these variables is not set, the corresponding command is determined by the {\tt PRINTER} environment variable (see \hyperref{Setting up a printer for \WAVE{}} {section~} {, page~\pageref{sec:printer-setup}} {sec:printer-setup}). You may change these commands (for example, to specify use of a different printer). When executing commands from its menu file, \WAVE{} replaces the strings {\tt \$PSPRINT} and {\tt \$TEXTPRINT} with the corresponding commands from the {\sf Print setup} window (see \hyperref{Menu Variables} {appendix~} {, page~\pageref{ch:menu-variables}} {ch:menu-variables}). \section{The {\sf Analyze} window} \index{Analyze window@{\sf Analyze} window} \begin{figure}[h] \centerline{\epsfig{file=analyze-window}} \end{figure} \label{sec:analyze-window} Open this window by selecting {\sf Analyze} from the \menubutton{File} menu. \begin{description} \item[\button{\tt <}] This button shifts the segment to be analyzed toward the beginning of the record, by an amount equal to the length of the segment if possible. \index{region of interest} \index{marker!{\tt <}} \index{elapsed time}\index{time!elapsed} \item[{\sf Start (elapsed)}] This field specifies (as elapsed time from the beginning of the record) the beginning of the segment to be analyzed. You may enter a time directly in this field, or you may insert a `{\tt <}' marker using the standard procedure for inserting annotations. \index{marker!{\tt >}} \item[{\sf End (elapsed)}] This field specifies the end of the segment to be analyzed. You may enter a time directly in this field, or you may insert a `{\tt >}' marker using the standard procedure for inserting annotations. \item[\button{\tt >}] This button shifts the segment to be analyzed toward the end of the record, by an amount equal to the length of the segment. \index{signal!number} \item[{\sf Signal}] This field specifies the selected signal. You may enter a signal number directly in this field, or you may point to a signal, depress the \keycap{Shift} key, and click left to select the signal. In the \WAVE{} menu file, the symbol `{\tt \$SIGNAL}' refers to the signal number of the selected signal; this symbol usually specifies a signal to be analyzed. The name of the selected signal appears to the right of its signal number. The uppermost signal displayed by \WAVE{} is signal 0. \index{absolute time}\index{time!absolute} \item[{\sf From}] This field specifies the absolute time (and date, if defined) of the beginning of the segment to be analyzed. The value in this field is updated automatically whenever the value in the {\sf Start (elapsed)} field changes, and vice versa. The {\sf From} field is disabled if the current record's header does not define the absolute time of the beginning of the record. \item[{\sf To}] This field specifies the absolute time (and date, if defined) of the end of the segment to be analyzed. The value in this field is updated automatically whenever the value in the {\sf End (elapsed)} field changes, and vice versa. The {\sf To} field is disabled if the current record's header does not define the absolute time of the beginning of the record. \index{signal!list} \item[{\sf Signal list}] This field specifies the signal list (a list of signal numbers, separated by spaces). In the \WAVE{} menu file, the symbol `{\tt \$SIGNALS}' refers to the signal list; this symbol usually appears where a list of signals to be analyzed is required. To change the signal list, either type into this field, or point to a signal, press and hold the \keycap{Control} key (to add the signal to the list) or the \keycap{$\Diamond$} (or \keycap{Alt}) key (to delete the first occurrence of the signal from the list), and click left. \index{Scope window@{\sf Scope} window} \item[\button{Show scope window}] This button pops up \WAVE{}'s \hyperref{{\sf Scope} window} {{\sf Scope} window (see section~} {, page~\pageref{sec:scope-window}} {sec:scope-window}, which can be used to display a signal in `oscilloscope' mode. \index{analysis commands window@{\sf Analysis Commands} window} \item[\button{Show command window}] This button pops up the {\sf Analysis Commands} window, a terminal emulator that receives commands generated by selecting most of the other buttons in this window, and that displays any text output of those commands. You may type commands directly into the window. \index{EDITOR environment variable}\index{environment variables!EDITOR} \index{textedit command@{\tt textedit} command}\index{menu file} \item[\button{Edit menu}] This button allows you to edit the \WAVE{} menu file, which configures the analysis buttons in the {\sf Analyze} window, using the text editor named in the {\tt EDITOR} environment variable (or {\tt textedit} if {\tt EDITOR} is not set). After you have saved your changes, select \button{Reread menu} to reconfigure this window. \item[\button{Reread menu}] Select this button to reconfigure the {\sf Analyze} window after you have made changes to the menu file (most easily done by using \button{Edit menu}) or after changing records. \item[\button{Reload}] This button causes \WAVE{} to reload the current annotation file. If a process is in progress in the {\sf Analysis Commands} window, \WAVE{} defers reloading until the process is finished. \end{description} \index{menu file} \index{action buttons} The remaining buttons in the {\sf Analyze} window perform actions determined by the \WAVE{} menu file. If the environment variable {\tt WAVEMENU} is set, it names that file; otherwise, the \WAVE{} menu file is {\tt wavemenu} (if it exists in the current directory) or {\tt /usr/lib/wavemenu.def}. \section{The {\sf Log} window} \index{Log window@{\sf Log} window} \begin{figure}[h] \centerline{\epsfig{file=log-window}} \end{figure} \label{sec:log-window} Open this window by selecting {\sf Log} from the \menubutton{File} menu. \index{log file} \begin{description} \item[{\sf File}] This field names the current log file. \item[\button{Load}] Press this button to load (or reload) the log file if an external process (such as one started from the {\sf Analyze} window, or an editor started using the \button{Edit} button in the {\sf Log} window) creates or modifies the log file. \item[{\sf Description}] This field contains the description associated with the current log entry. \item[{\sf Delay}] This slider controls the interval between display updates when a log review is in progress. \index{marker!{\tt :}} \item[\button{Add}] This button adds the current record name, the time corresponding to the center of the signal window, and the contents of the description field, as an entry in the log file; it also inserts an index mark (`{\tt :}') in the center of the signal window. \item[\button{Replace}] This button replaces the description attached to the current log entry with the current contents of the description field. It does not create a new entry (use the \button{Add} button for that purpose). \item[\button{Delete}] This button deletes the current entry from the log. This button also causes \WAVE{} to display the next log entry if it exists. \index{EDITOR environment variable}\index{environment variables!EDITOR} \item[\button{Edit}] This button opens the current log file using the text editor named in the \index{environment variables!EDITOR}\index{EDITOR environment variable} \index{textedit command@{\tt textedit} command} {\tt EDITOR} environment variable (or {\tt textedit} if {\tt EDITOR} is not set). Save your edits, exit from the editor, and click on the \button{Load} button in the {\sf Log} window before attempting to make changes to the log using \button{Add}, \button{Replace}, or \button{Delete}. \item[\button{\tt |<}] This button causes \WAVE{} to ``rewind'' the log (i.e., to show the first log entry). \item[\button{\tt <<}] This button causes \WAVE{} to begin reviewing each entry in the log file in reverse order, pausing 5 seconds between entries. While a review is in progress, only the \button{Pause} button is enabled. \item[\button{\tt <}] This button causes \WAVE{} to show the previous log entry. \item[\button{Pause}] This button causes \WAVE{} to stop the review of the log file that was begun by \button{\tt <<} or \button{\tt >>}. The \button{Pause} button is disabled unless a review is in progress. \item[\button{\tt >}] This button causes \WAVE{} to show the next log entry. \index{log file!review} \item[\button{\tt >>}] This button causes \WAVE{} to begin reviewing each entry in the log file, pausing 5 seconds between entries. While a review is in progress, only the \button{Pause} button is enabled. \item[\button{\tt >|}] This button causes \WAVE{} to ``fast forward'' the log (i.e., to show the last log entry). \end{description} \section{The {\sf View} window} \index{View window@{\sf View} window} \begin{figure}[h] \centerline{\epsfig{file=view-window}} \end{figure} \label{sec:view-window} Open this window using \button{View...} in \WAVE{}'s main window. \index{annotation!mnemonic} \index{annotation!subtyp@{\tt subtyp}} \index{subtyp in annotation@{\tt subtyp} in annotation} \index{annotation!chan@{\tt chan}} \index{chan in annotation@{\tt chan} in annotation} \index{annotation!num@{\tt num}} \index{num in annotation@{\tt num} in annotation} \index{annotation!aux@{\tt aux}} \index{aux in annotation@{\tt aux} in annotation} \index{signal!names} \index{signal!baselines} \index{baselines} \index{marker bars} \index{signal!levels} \begin{description} \item[{\sf Show}] Toggle these options by selecting them. Multiple annotation fields are shown in the following arrangement: \begin{quote} {\it annotation mnemonic}\\ {\it subtype}\\ {\tt chan} {\it field}\\ {\tt num} {\it field}\\ {\tt aux} {\it field} \end{quote} {\sf Signal names} and {\sf baselines} are defined in the header file for the current record. {\sf Markers} show the precise locations of all annotations. {\sf Levels} show the amplitudes of each signal at the time indicated by the pointer, whenever a mouse button is depressed. The {\sf Level} window appears at this time. \index{time scale}\index{scales!time} \item[\amenubutton{Time scale:}] Set the horizontal scale for the signal and {\sf Scope} windows by selecting one of the choices. \index{scales!amplitude} \index{amplitude scale} \item[\amenubutton{Amplitude scale:}] Set the vertical scale for the signal and {\sf Scope} windows by selecting one of the choices. \end{description} \index{grid}\index{resolution!of display} \index{display calibration}\index{calibration!display} \index{dpi option for WAVE@{\tt -dpi} option for \WAVE{}} \index{command-line options!{\tt -dpi}}\index{options!{\tt -dpi}} You can check the calibration of the display by enabling the 0.2 s x 0.5 mV grid, and then by measuring the spacing of the grid lines. If the spacing is incorrect, your X server does not know the actual display resolution. See your X server documentation if this is the case, or use the {\tt -dpi} option when starting \WAVE{} (start \WAVE{} with no arguments for instructions). \begin{description} \index{signal!display mode} \index{signal!number} \index{signal!list} \item[\amenubutton{Draw:}] Use this control to toggle \WAVE{}'s signal display mode. By default, \WAVE{} displays all signals in order of signal number, with signal 0 at the top of the signal window. If you select {\sf listed signals only}, \WAVE{} displays only those signals that appear in the signal list (in the {\sf Analyze} window, from top to bottom in the order in which they appear in the signal list). \index{format!of annotation display} \index{annotation!display mode} \index{chan in annotation@{\tt chan} in annotation} \index{num in annotation@{\tt num} in annotation} \item[\amenubutton{Show annotations:}] Use these menu buttons to choose how annotations are to be displayed. By default, \WAVE{} displays annotations in the center of the signal window. If you select {\sf attached to signals}, each annotation appears near the signal specified by its {\tt chan} field. If you select {\sf as a signal}, \WAVE{} draws a signal derived from the {\tt num} fields of any annotations in the window, in place of the standard annotation display. The right-hand menu button allows you to choose if the full texts of closely-spaced annotations should be shown even if they overlap (if you choose {\sf allow overlap}). By default, the texts are truncated after the first character if they would overlap, to improve legibility. \index{format!of time display/entry} \index{time!display mode} \index{absolute time} \index{elapsed time} \item[\amenubutton{Time display:}] By default, \WAVE{} displays elapsed time from the beginning of the record in {\sf hh:mm:ss} format. Use this control to select display of absolute time (if defined for the record), or to display time in sample intervals. If you select absolute time display, you may enter absolute times in the {\sf Find} window. \index{format!of grid display} \index{grid} \item[\amenubutton{Grid:}] Use this menu button to select a grid style, or to suppress the grid display. \item[\button{Undo changes}] This button cancels any changes you have made to the {\sf View} window settings and restores the indicators to reflect the current settings. \item[\button{Redraw}] This button causes \WAVE{} to accept any changes you have made in the {\sf View} window. Pressing \button{Redraw} dismisses the {\sf View} window and refreshes the signal window. \index{Xdefaults file@{\tt .Xdefaults} file} \item[\button{Save as new defaults}] This button preserves the current settings within the {\sf View} window in your home {\tt .Xdefaults} file. \end{description} \section{The {\sf Edit} menu} \index{Edit menu@{\sf Edit} menu} \label{sec:edit-menu} Open this menu using \menubutton{Edit} in \WAVE{}'s main window. \begin{description} \item[{\sf Allow editing}] This selection allows you to make changes to the annotation buffer. \index{marker!editing} \item[{\sf View only}] This selection disallows annotation editing. You may still edit `{\tt <}', `{\sf :}', and `{\tt >}' markers. \end{description} \section{The {\sf Properties} menu} \index{Properties menu@{\sf Properties} menu} \label{sec:prop-menu} Open this menu using \menubutton{Properties} in \WAVE{}'s main window. \index{wfdbdesc command@{\tt wfdbdesc} command} \begin{description} \item[{\sf Signals...}] This selection pops up a window containing information about the signals of the current record, obtained by running \htmladdnormallink{{\tt wfdbdesc}}{../wag/wfdbde-1.htm}. \index{sumann command@{\tt sumann} command} \item[{\sf Annotations...}] This selection pops up a window containing a summary of the contents of the annotation buffer, obtained by running \htmladdnormallink{{\tt sumann}}{../wag/sumann-1.htm} (after saving any edits). \index{WAVE version number@\WAVE{} version number} \item[{\sf About Wave...}] This selection pops up a window containing the version number and date of this copy of \WAVE{}. This window may also contain news about recent changes in \WAVE{}. \end{description} \section{The {\sf Find} window} \index{Find window@{\sf Find} window} \begin{figure}[h] \centerline{\epsfig{file=find-window}} \end{figure} \label{sec:find-window} Open this window using \button{Find...} in \WAVE{}'s main window. \begin{description} \index{View window@{\sf View} window} \item[{\sf Start time}] Specifies the time of the sample shown at the left edge of the signal window, in the format specified by the {\sf Time display} item in the {\sf View} window. Go to any other part of the record by entering the time in this field. \item[{\sf End time}] Specifies the time of the sample shown at the right edge of the signal window, in the format specified by the {\sf Time display} item in the {\sf View} window. Go to any other part of the record by entering the time in this field. \end{description} \index{format!of time display/entry} Times can be entered in {\tt h:m:s} format, with hours or hours and minutes omitted, or in {\tt s{\it nnnnn}} format, in which {\it nnnnn} is a number of sample intervals from the beginning of the record. \index{annotation!mnemonic} \begin{description} \index{search}\index{annotation!searching for} \item[{\sf Search for annotation}] Specifies a target for \button{{\tt <} Search} and \button{Search {\tt >}}. Changing this field causes an immediate forward search. The contents of this field should match an annotation or marker mnemonic, signal quality code, rhythm, comment, or other text string, or one of the following: \begin{tabular*}{\textwidth}{l p{3.5 in}} {\tt *v} & {matches any ventricular ectopic beat} \\ {\tt *s} & {matches any supraventricular ectopic beat} \\ {\tt *n} & {matches any other beat type} \\ {\tt *} & {matches any annotation or marker} \\ {\tt .} & {matches a deletion made during this \WAVE{} session} \\ \end{tabular*} \index{search template window@{\sf Search Template} window} \item[\button{More options...}] This button clears the contents of the {\sf Search for} field and opens the \hyperref{{\sf Search Template} window} {{\sf Search Template} window (see section~} {, page~\pageref{sec:search-template-window})} {sec:search-template-window}. \index{signal!searching for}\index{search} \item[{\sf Find signal}] Specifies the name of a signal. Changing this field causes an immediate forward search for the next valid samples of the specified signal. To repeat the search, select the field and press \keycap{Enter}. \end{description} \section{The {\sf Help Topics} window} \index{on-line help}\index{Help Topics window@{\sf Help Topics} window} \begin{figure}[h] \centerline{\epsfig{file=help-topics}} \end{figure} \label{sec:help-topics-window} Open this window using \button{Help} in \WAVE{}'s main window. \index{help!printing} Each of the buttons in this window, except for the last two described below, opens a help file in a scrollable text window. Use the \button{Print} button in each window to get a paper copy of the help file if you wish. The on-line manual for \WAVE{} includes many of these files verbatim. \begin{description} \item[\button{Introduction}] A quick introduction to \WAVE{}'s features. \item[\button{Buttons}] A summary of the functions of the buttons in \WAVE{}'s main control panel and pop-up windows. \item[\button{Annotation Editing}] A summary of annotation editing procedures. \item[\button{WAVE Logs}] How to create and review \WAVE{} log files. \item[\button{Printing}] Methods for printing annotated signals from \WAVE{}. \item[\button{Analysis}] Using and customizing the {\sf Analyze} window. \item[\button{Resources}] \WAVE{}-specific X11 resources. \item[\button{What's new}] Recent changes in \WAVE{}. \item[\button{Frequently asked questions}] Common questions about \WAVE{} (answers, too!). \item[\button{User's Guide}] Click on this button to begin reading this guide in a web browser window. By default, \WAVE{} uses Mozilla. To configure \WAVE{} to use a different browser, see \hyperref{\WAVE{} and the web} {section~} {, page~\pageref{sec:web}} {sec:web}. \index{Mozilla}\index{web browser} \item[\button{Quit from Help}] Press this button to dismiss the {\sf Help Topics} window. \end{description} \section{The {\sf Annotation Template} window} \index{Annotation Template window@{\sf Annotation Template} window} \begin{figure}[h] \centerline{\epsfig{file=annotation-template}} \end{figure} \label{sec:annotation-template-window} Open this window at any time by clicking left anywhere within the signal window. Use the data fields in the {\sf Annotation Template} before inserting an annotation, to specify the characteristics of the annotation you wish to insert. \begin{description} \item[\amenubutton{Type:}] This field specifies the type of annotation to be inserted. It may be changed by selecting a new value from the pull-down menu, by typing the mnemonic while the pointer is within the signal window, or by selecting an existing annotation and pressing the \keycap{Copy} or \keycap{F6} keys (these key commands also copy the other fields of the selected annotation into the corresponding fields of the {\sf Annotation Template}). \item[{\sf Text}] This field specifies the contents of the optional annotation {\tt aux} field. It may be changed by typing into it directly or by selecting an existing annotation and pressing the \keycap{Copy} or \keycap{F6} keys. In most cases, it should be empty, but it must be filled in for rhythm and certain other non-beat annotations. When the annotation is written, \WAVE{} prefixes the required byte count to this field before transferring it to the {\tt aux} field. \item[{\sf Subtype}] This field specifies the contents of the annotation {\tt subtyp} field. In most cases, it should be 0; legal values range from -128 to +127. \item[{\sf `Chan' field}] This field specifies the contents of the annotation {\tt chan} field. In most cases, it should be 0; legal values range from -128 to +127. In multi-edit mode, the {\tt chan} field of the annotation indicates the signal number of the attached signal. When inserting annotations in multi-edit mode, the value of the {\tt chan} field in the annotation is determined by which signal is nearest to the pointer when the insertion is performed, and the {\sf `Chan' field} in the {\sf Annotation Template} is updated accordingly. \item[{\sf `Num' field}] This field specifies the contents of the annotation {\tt num} field. In most cases, it should be 0; legal values range from -128 to +127. If you have chosen to {\sf Show annotations: as a signal} in the {\sf View} window, the {\tt num} fields of the annotations determine the signal amplitudes. \item[\button{Change all in range}] This button changes all annotations between the `{\tt <}' and `{\tt >}' markers (the {\sf Start} and {\sf End} times in the {\sf Analyze} window) to match the {\sf Annotation Template}. \item[\button{Dismiss}] This button makes the {\sf Annotation Template} window disappear (until it is recalled by clicking left while the pointer is within the signal window). \end{description} \section{The {\sf Search Template} window} \index{search template window@{\sf Search Template} window} \begin{figure}[h] \centerline{\epsfig{file=search-template}} \end{figure} \label{sec:search-template-window} Open this window by clicking left on \button{More Options...} in the {\sf Find} window. This window specifies criteria for searches performed using \button{{\tt <} Search} and \button{Search {\tt >}}. Each of the five menu buttons at the left edge can be set to {\sf Ignore} (the default) or {\sf Match} in order to exclude or include, respectively, the associated annotation field in the search criteria. The {\sf Type}, {\sf Text}, {\sf Subtype}, {\sf `Chan' field}, and {\sf `Num' field} controls work in the same way as their counterparts in the \hyperref{{\sf Annotation Template} window} {{\sf Annotation Template} window (see section~} {, page~\pageref{sec:annotation-template-window})} {sec:annotation-template-window}. \begin{description} \item[\button{Match selected annotation}] This button copies the fields of the selected annotation (if any) into the {\sf Search Template}, and sets all five of the menu buttons at the left edge of the window to {\sf Match}. This is a convenient shortcut if you wish to find another annotation identical to a given one. \item[\button{Dismiss}] This button makes the {\sf Search Template} window disappear (until it is recalled by clicking left on \button{More Options...} in the {\sf Find} window). \end{description} \section{The {\sf Level} window} \index{Level window@{\sf Level} window} \begin{figure}[h] \centerline{\epsfig{file=level-window}} \end{figure} \label{sec:level-window} Open the {\sf Level} window by choosing {\sf level} in the {\sf View} window, then by clicking a mouse button anywhere in the signal window. \begin{description} \item[\amenubutton{Show}] Choose how time and signal levels are displayed in the {\sf Level} window using this menu. The four menu choices are {\sf physical units (absolute)}, {\sf physical units (relative)}, {\sf WFDB units (absolute)}, and {\sf WFDB units (relative)}. Physical units are seconds for time, and as specified in the record's header file for each of the signals. WFDB units are sample intervals and analog-to-digital units (adu). In relative mode, all measurements are shown as differences beween the current location and the reference location (which is determined by the placement of the `{\tt ;'} reference marker). \item[{\sf Time}] This item shows the time corresponding to the most recent cursor position while a button was depressed within the signal window (and to the signal levels shown in the remainder of the {\sf Level} window). Depending on the choice made using the {\sf Show} menu, this item may be named {\sf Interval} or {\sf Sample number}. \item[{\it Levels}] Most of the {\sf Level} window contains an informational display indicating the names and levels (amplitudes) of the signals. If an asterisk (*) appears at the end of any line, it indicates that the corresponding signal is uncalibrated. \item[\button{Dismiss}] This button makes the {\sf Level} window disappear (until it is recalled by clicking within the signal window). \end{description} \section{The {\sf Scope} window} \index{Scope window@{\sf Scope} window} \index{Analyze window@{\sf Analyze} window} \index{oscilloscope display} \begin{wrapfigure}{l}{2cm} \mbox{\epsfig{file=scope-controls}} \end{wrapfigure} \label{sec:scope-window} Open the {\sf Scope} window from the {\sf Analyze} window, using \button{Show scope window}. The {\sf Scope} window provides a simulated oscilloscope display of one signal, triggered by QRS (beat) annotations. (Annotations that are not considered QRS annotations, such as rhythm changes and comments, are those that appear displaced vertically with respect to the QRS annotations in the signal window; these non-QRS annotations do not trigger the scope display.) \index{scales!time}\index{scales!amplitude} \index{time scale}\index{amplitude scale} Select the signal to be drawn in the {\sf Scope} window using the {\sf Signal} control in the {\sf Analyze} window. The \amenubutton{Time scale:} and \amenubutton{Amplitude scale} controls in the {\sf View} window define the scales used in the {\sf Scope} window. \begin{description} \item[{\sf Speed}] This slider controls the playback speed. Click left on the slider and drag it to the desired position -- lower to reduce speed, higher to increase speed. \item[{\sf dt}] This field specifies the interval from the left edge of the {\sf Scope} window to the annotated sample, in seconds. Negative values may be used to shift the left edge of the scope window past the annotated sample. Drag the resize corners to change the interval between the left and right edges of the scope window. \item[\button{\tt <<}] This button displays frames continuously in reverse order. Use \button{Pause} to interrupt the display. The display stops when it reaches the beginning of the record, an index mark (`{\tt :}'), or the `{\tt <}' marker. \item[\button{\tt <}] This button displays the previous frame. \item[\button{Pause}] This button freezes the scope display. It also forces the signal window to be redrawn, roughly centered on the time shown at the bottom of the scope window, with the current annotation selected. \item[\button{\tt >}] This button displays the next frame. \item[\button{\tt >>}] This button displays frames continuously in normal order. Use \button{Pause} to interrupt the display. The display stops when it reaches the end of the record, an index mark (`{\tt :}'), or the `{\tt >}' marker. \end{description} \chapter{System Requirements for \WAVE{}} \label{app:system-requirements} \index{Linux}\index{Solaris}\index{SunOS}\index{SPARCstation}\index{PC} \index{Macintosh}\index{Microsoft Windows} This section includes a shopping list of hardware and software that you will need in order to run \WAVE{}, or that you may find useful in order to use \WAVE{} most effectively. \section{Necessities} At a minimum, you will require: \begin{description} \item [\WAVE{} software] This is freely available from PhysioNet in source form, and in binary form for PCs running GNU/Linux or MS-Windows, or Macintoshes running OS X. (Older binaries for SPARC-based systems running Solaris 2.x or SunOS 4.1.x are also available on PhysioNet. See \hyperref{how to obtain the current version of \WAVE{}} {section~} {, page~\pageref{sec:getting-wave} for details} {sec:getting-wave}.) If the XView toolkit can be installed on your platform, it should be easy to build and use \WAVE{} on it. \index{WAVE host@\WAVE{} host} \index{resolution!of display} \index{X Window System!XFree86 server} \item [A computer capable of acting as a \WAVE{} host] Virtually any PC with a 386 or better CPU can run Linux, and such systems are likely to be the least expensive choice. Ideally, a Linux PC to be used as a \WAVE{} host should have at least 8 Mb of RAM, at least 200 Mb of available disk space, a three-button mouse (or trackball), and a graphics card and monitor (17-inch or larger, with a dot pitch of .26 mm or less) capable of non-interlaced display at 65 Hz or faster with a resolution of at least 1024x768 with 256 colors. In most cases, you will also want the system to be equipped with an Internet connection (for obtaining data and software from PhysioNet and other sources).. Most PCs manufactured since 1995 will easily meet these requirements; new PCs will exceed most of them by large factors. In mid-2005, it was possible to assemble a suitable Linux PC for about US\$200 (not including the monitor). It is not unreasonable to budget an equal or greater amount for a good monitor, since \WAVE{}'s usability depends to a significant extent on being able to see its output clearly. If your budget permits, a flat-panel (LCD) monitor is an excellent choice, particularly if you plan to do much annotation editing, because these monitors typically present very stable images that do not tire the eye. Fully configured and supported Linux PCs are available from many sources if you prefer not to assemble your own; try a Google search for ``linux pc'' to find vendors. Inexpensive three-button mice, trackballs, and touchpads manufactured by Logitech and many others are widely available for PCs, and are highly recommended if you do much annotation editing. Most are fully compatible with Microsoft two-button mice. Some users prefer trackballs for precision editing since there is no tendency for the pointer to move when clicking the buttons, as with a mouse. Future versions of \WAVE{} will make use of scroll wheels on mice when available; if you choose a mouse equipped with a scroll wheel, try to select one that allows you to use the middle button without generating scroll wheel events. For those on a tight budget, suitable used PCs are often available for next to nothing. If you use an older PC and can afford to upgrade components, get a new three-button mouse or trackball first, a new monitor if the old one is inadequate, more memory if you have less than 64 Mb, a new IDE disk drive (around US\$100) if yours is more than 3 years old, and only then consider other performance upgrades such as a faster CPU. Although Linux does not require large amounts of RAM, it can use additional RAM very effectively, and you are likely to find that purchasing (say) 16 or 32 Mb of additional RAM results in a bigger performance improvement than spending the same amount on a faster CPU. Although \WAVE{} now runs on MS-Windows, we still recommend using GNU/Linux as the platform of choice because of its greater speed, capability, configurability, security, and stability, and because of the wider range of related software available for GNU/Linux. Other \WAVE{} users have reported success on PCs running FreeBSD. Macintoshes running OS X and SPARC-based systems may also be worth considering, although they tend to be substantially more expensive than comparable or faster Linux or FreeBSD PCs. \end{description} \section{Printers} \index{printing}\index{PostScript}\index{Ghostscript} All currently available \WAVE{}-compatible applications for printing require the use of a PostScript (or compatible) printer accessible to the \WAVE{} host \index{WAVE host@\WAVE{} host} system. If you need paper output, therefore, you should obtain a PostScript printer, or a PostScript-compatible interpreter (such as the freely available GNU Ghostscript) for your printer, or you should be prepared to write your own printing applications. \section{Remote access requirements} \index{network access}\index{modem access}\index{PPP}\index{SLIP}\index{TERM} \index{Ethernet} Since any recently manufactured PC or Macintosh can act as a \WAVE{} host, remote access to \WAVE{} is likely to be of interest mainly to those seeking to use older hardware. For this purpose, virtually any local system (computer or X terminal) connected by TCP/IP-based Ethernet to a \WAVE{} host can be used. (PPP, SLIP, or TERM serial-line connections also work, and may be acceptable for viewing data, but even fast serial connections are likely to be intolerably slow for extensive annotation editing.) Thus it is possible for several users to access \WAVE{} simultaneously using a single \WAVE{} host, provided only that each user's computer or terminal runs an X server. Free or inexpensive X servers are available for virtually any computer made in the last 20 years, including those running any version of UNIX, as well as for MS-DOS and MS-Windows PCs, Macintoshes, Amigas, and DEC VAXen running VMS. \section{Data} \index{MIT-BIH Arrhythmia Database} \index{European ST-T Database}\index{MIT-BIH Polysomnographic Database} \index{MGH/MF Waveform Database} You may find that an existing database of digitally recorded signals may be useful for your studies. PhysioNet, at \htmladdnormallink{http://physionet.org/,}{http:/physionet.org/}, offers free on-line access to over 40 such databases with thousands of recordings ranging in length from a few seconds to several days. These include all or most of the MIT-BIH Arrhythmia Database, the European ST-T Database, the MIT-BIH Polysomnographic Database, the MGH/MF Waveform Database, and the Long-Term ST Database (which are also available on CD-ROMs from their creators), and many other databases available only via PhysioNet. \WAVE{} can read these databases directly from PhysioNet without downloading them first (the WFDB library, which handles reading and writing data for \WAVE{}, contains HTTP client code that permits applications such as \WAVE{} to read input with equal ease from local disks or remote web servers). This is an excellent way to explore PhysioBank (PhysioNet's collection of signal databases). \section{About Linux} \label{sec:linux} \index{Linux} As mentioned earlier, an excellent choice for a \WAVE{} host is a PC running Linux. Linux is a (very) complete, and completely free, robust, modern reimplementation of the UNIX operating system, written by Linus Torvalds and a cast of thousands. It is freely available in source and ready-to-run form from many sources (LWN.net maintains a list of over 400 sites at \htmladdnormallink{http://lwn.net/Distributions/}{http://lwn.net/Distributions/}). You can also obtain Linux on CD-ROMs from many commercial sources, generally at very low prices (typically US\$10 to \$40, depending mainly on the amount of printed documentation and technical support offered). Current Linux distributions include TCP/IP networking including NFS support, the complete collection of GNU software including the GNU C/C++ compiler, Ghostscript, \TeX, X11R6, and much more. Some also include \index{XView}\index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} XView 3.2, {\tt olwm} and {\tt olvwm}; these are also available from PhysioNet. For further information, visit the home page of the Linux Documentation Project (\htmladdnormallink{{\tt http://www.tldp.org/}} {http://www.tldp.org/}), where you may also find the Linux HOWTO documents mentioned earlier. \section{Other useful software} \label{sec:other-software} \index{WFDB Software Package} \WAVE{} makes extensive use of other components of the \emph{WFDB Software Package} of which it is a part. The most recent version of this package is always freely available from PhysioNet (\htmladdnormallink{{\tt http://physionet.org/}} {http://physionet.org/}). The WFDB Software Package includes {\tt calsig}, {\tt mrgann}, {\tt plot2d}, {\tt pschart}, {\tt psfd}, {\tt rdann}, {\tt rdsamp}, {\tt sample}, {\tt snip}, {\tt sqrs}, {\tt tach}, {\tt wfdbcollate}, {\tt wfdbdesc}, {\tt wfdbwhich}, {\tt wrann}, {\tt wrsamp}, and {\tt xform} (among many other applications). The package also includes the WFDB library of interface functions for user-written applications that read and write signal and annotation files in the formats supported by \WAVE{}. A web browser, though not a necessity for everyone who uses \WAVE{}, should be part of your software toolbox. If you use link annotations, you will need a browser in order to follow the links to the external data. Even if you don't anticipate using link annotations, you can still use a web browser to view the on-line version of this guide. Mozilla is an attractive choice because of its simple remote-control interface, among other reasons. Mozilla may be obtained freely from \htmladdnormallink{{\tt http://\-www\-.mozilla\-.org}} {http://www.mozilla.org/}). \index{Galeon}\index{Konqueror}\index{Mozilla}\index{Netscape}\index{Opera} \index{web browser} \index{gnuplot command@{\tt gnuplot} command} \index{plot2d command@{\tt plot2d} command} \index{plt command@{\tt plt} command} An X-Y plotting program capable of being run from the command line is another nearly essential tool. A suitable program should be able to accept multicolumn text input, allowing specification of which columns to plot from the command line (so that it can be driven by \WAVE{}). Avoid commercial software designed for business graphics with arbitrary (often very low) limits on the number of points that can be plotted. If you expect to prepare plots for publication, avoid packages that only provide screen dumps; any decent plotting program should be able to generate plots at the resolution of your printer. \htmladdnormallink{{\tt plt}}{../wag/plt-1.htm} is a highly capable plotting package freely available from \htmladdnormallink{{\tt physionet.org};}{http://physionet.org/}. \htmladdnormallink{{\tt plot2d}}{../wag/plot2d-1.htm} (included in the WFDB Software Package) is a bare-bones command-line front end to {\tt gnuplot}, a fairly capable interactive plotting program that is included in most Linux distributions, and is also available from \htmladdnormallink{http://www.gnuplot.info/}{http://www.gnuplot.info/} and numerous other sites. {\tt plot2d} uses {\tt plt} rather than {\tt gnuplot} if {\tt plt} is installed. \chapter{Setting up a \WAVE{}~host} \label{app:setup} \section{Obtaining and installing the current version of \WAVE{}} \label{sec:getting-wave} \index{WAVE host@\WAVE{} host} \index{Linux}\index{Solaris}\index{SunOS}\index{SPARCstation}\index{PC} \index{WWW}\index{WFDB Software Package}\index{downloading} \index{installing \WAVE{}} Current sources for \WAVE{}, and precompiled \WAVE{} binaries for GNU/Linux or Mac OS X may be obtained from PhysioNet. Point your Web browser to \htmladdnormallink{{\tt http://physionet.org/}}{http://physionet.org/} for details. Note that \WAVE{} cannot be run at all until the WFDB library (contained within the WFDB Software Package) and the XView libraries ({\tt libxview.so.3} and {\tt libolgx.so.3}) have been installed. It is strongly recommended that you install the complete WFDB Software Package, since \WAVE{} uses many of the applications included in this package. The W3C's {\tt libwww} libraries, or the newer and more capable {\tt libcurl} library, are optional but recommended; either {\tt libwww} or {\tt libcurl} is necessary if you wish to use \WAVE{} to read data directly from web and FTP servers. \index{XView} \index{Linux} If you would like to try porting \WAVE{} to another operating system for which X11 client support (Xlib) is available, please obtain the XView source distribution and attempt to port `{\tt cmdtool}' (a simple terminal emulator included in the XView distribution) first. Compared to the original sources, the Linux/Mac OS X/MS-Windows version of XView (freely available from \htmladdnormallink{PhysioNet}{http://physionet.org/physiotools/xview/}, \htmladdnormallink{{\tt metalab.unc.edu}} {ftp://metalab.unc.edu/pub/Linux/libs/X/xview/}, \htmladdnormallink{{\tt www.rpmfind.net}} {http://www.rpmfind.net/linux/RPM/}, and their mirrors, and also available on many low-cost Linux CD-ROM archives) is easier to port to another operating system, since many Sun-specific dependencies were removed. \section{Setting up a printer for \WAVE{}} \label{sec:printer-setup} \index{lpr command@{\tt lpr} command} \index{PostScript} \index{Solaris} \WAVE{}'s printing capabilities (see \hyperref{Printing} {chapter~} {, page~\pageref{ch:printing}} {ch:printing}) require that the \WAVE{} host be able to print PostScript. Your \WAVE{} host may already be set up to do this if its default printer is a PostScript printer. (Under Solaris, make sure that the directory containing the BSD command {\tt lpr} is in your {\tt PATH}.) \index{Ghostscript} \index{PostScript} \index{printer!non-PostScript} \index{fax output} \index{format!of graphics files} If the printer itself is incapable of rendering PostScript, you may still be able to use it if it is supported by Ghostscript (a freely available PostScript interpreter). Ghostscript is included in executable and portable C source form in most if not all Linux distributions, and is also available from many sources, including its own \htmladdnormallink{web site}{http://www.cs.wisc.edu/~ghost/} and its mirrors, \htmladdnormallink{CTAN}{http://ctan.tug.org}, and other archives of GNU software. Ghostscript supports a wide variety of ink jet, dot matrix, and laser printers including popular models made by Canon, Epson, Hewlett Packard, IBM, and others; it can also render PostScript into files in a wide variety of graphics formats, including PBM, PGM, PPM, PCX, and TIFF (including G3 fax format -- with a fax modem, you can even print PostScript output on any fax machine). Ghostscript works by rasterizing Postscript input on the host system, then transmitting the raster image to the printer in whatever format the printer accepts. For details on setting up Ghostscript for your printer, see the documentation that comes with Ghostscript, or the Linux Printing-HOWTO. Assuming that the \WAVE{} host is somehow able to print PostScript output (either directly to a PostScript printer, or via Ghostscript or another interpreter), setup for \WAVE{} is straightforward: \begin{itemize} \index{PRINTER environment variable}\index{environment variables!PRINTER} \item Set the {\tt PRINTER} environment variable to the name of the printer you wish to use, if that printer is not the default printer on the \WAVE{} host. \index{PSPRINT environment variable}\index{environment variables!PSPRINT} \index{lpr command@{\tt lpr} command} \item Set the {\tt PSPRINT} environment variable to the command needed to print PostScript from the standard input on the desired printer. \WAVE{} assumes that this command is `{\tt lpr}' if neither {\tt PSPRINT} nor {\tt PRINTER} are set; if only {\tt PRINTER} is set, \WAVE{} assumes that this command is `{\tt lpr -P\$PRINTER}'. \index{TEXTPRINT environment variable}\index{environment variables!TEXTPRINT} \index{on-line help}\index{help!printing} \item Set the {\tt TEXTPRINT} environment variable to the command needed to print ordinary text from the standard input on the desired printer. (This command is used, for example, to print the text of help topics.) \WAVE{} assumes that this command is `{\tt lpr}' if neither {\tt TEXTPRINT} nor {\tt PRINTER} are set; if only {\tt PRINTER} is set, \WAVE{} assumes that this command is `{\tt lpr -P\$PRINTER}'. \index{printer!resolution} \index{resolution!of printed output} \item Finally, if your printer is capable of resolution higher than 300 dpi, you may also wish to add appropriate {\tt -d} options to the {\tt pschart} and {\tt psfd} commands in \WAVE{}'s menu file (see \hyperref{Creating custom formats for printing} {section~} {, page~\pageref{sec:custom-printing}, for details} {sec:custom-printing}). (This is not necessary in order to obtain properly-scaled output, but this will affect the amount of fine detail in your plots, and the thickness of the plotting lines.) \end{itemize} \index{profile@{\tt .profile}} \index{login@{\tt .login}} In most cases, you will want to add the commands needed to set these variables to your {\tt .login} or {\tt .profile} file on the \WAVE{} host, so that they are executed automatically whenever you log in. For details on setting environment variables, see the discussion of the \hyperref{{\tt DISPLAY} variable} {{\tt DISPLAY} variable in section~} {, page~\pageref{sec:setting-display}} {sec:setting-display}. \index{printing!remotely vs. locally} If you are using \WAVE{} remotely, you may prefer to use a local printer for your output. \WAVE{} has no specific support for doing so, but this can often be arranged if, for example, your own computer is able to make its printer available for printing from the \WAVE{} host (this is can be done easily if your computer runs any version of UNIX, or with somewhat more difficulty if it runs Microsoft Windows). In this case, usually all that is needed is to set {\tt PRINTER} to the name of your printer (as it is known to the \WAVE{} host). Another possibility is to set {\tt PSPRINT} and {\tt TEXTPRINT} to commands that capture output in a file, and then transfer the file to your own computer (via FTP or any other method) for local printing. For example, you might set {\tt PSPRINT} to `{\tt cat~>>output.ps}' and {\tt TEXTPRINT} to `{\tt cat~>>output.txt}'. These commands accumulate all PostScript output into {\tt output.ps}, and all text output into {\tt output.txt}, so you should delete these files from the \WAVE{} host after copying them to your computer, so they don't grow indefinitely. If you need to change printers or print commands while running \WAVE{}, this can be easily accomplished within \WAVE{}'s \hyperref{{\sf Print setup} window} {{\sf Print setup} window (see section~} {, page~\pageref{sec:print-setup-window})} {sec:print-setup-window}, which may be opened by selecting {\sf Print setup} from the \menubutton{File} menu. \chapter{Frequently Asked Questions} This is the \WAVE{} FAQ (frequently asked questions) list. If your version of \WAVE{} is newer than this guide, you may wish to review the on-line copy of this FAQ (click on \button{Frequently Asked Questions} in \WAVE{}'s {\sf Help} window). If your question is not on this list, and you think it should be, please send it to me (\htmladdnormallink{{\tt wfdb@physionet.org}} {mailto:wfdb@physionet.org})! \section{Hardware questions} \subsection{Can I use \WAVE{} if I don't run Linux or another Unix?} Why would you not want to run Linux or Unix? Any version of Linux or Unix, including Mac OS X, is a much better choice for research than any version of MS-Windows. See \htmladdnormallink{{\tt http://\-srom.zgp.org/}}{http://srom.zgp.org/} for an independent perspective on this issue. Linux can coexist on the same PC with MS-Windows if necessary. The major obstacle to porting \WAVE{} to other platforms is that the XView toolkit is needed for the user interface components of \WAVE{}. XView is free software and can be ported to other platforms, but this task is not trivial. Since late 2004 it has been possible to run \WAVE{} on MS-Windows directly, using the free Cygwin POSIX emulation library and X11 server. If you have any networked computer that can run X11R4 or a later version (this includes all current UNIX workstations, PCs, Macintoshes, and a variety of other systems), and access via network to another computer that can host \WAVE{} (see above), you can run \WAVE{} remotely (see the next question). \subsection{How can I use \WAVE{} from an X terminal, PC, Mac, etc.?} \index{remote access}\index{network access} As with any X application, you can use an X terminal or any system equipped with an X server to interact with a copy of \WAVE{} running on another computer connected to the same network. To do so, however, requires a few additional steps: \begin{itemize} \item Find out the host names of the two systems. On a UNIX system, the `{\tt hostname}' command will give you this information. If your display is connected to a PC or other non-UNIX system, you can try remotely logging onto the UNIX system on which \WAVE{} resides, and using the `{\tt who}' command to discover the name of the system in front of you. \item Assume that your system is named {\tt hither} and that \WAVE{} resides on {\tt yon}. \index{WAVE host@\WAVE{} host} \index{IP address} \index{address, IP (Internet Protocol)} You may need to add {\tt yon} to the X server access control list on {\tt hither}, in order to make it possible for X clients on {\tt yon} to open windows on your screen. If {\tt hither} is a UNIX system, you can accomplish this by typing (on {\tt hither}) `{\tt xhost +yon}'. (Note: if {\tt hither} doesn't recognize {\tt yon} by name, this command will have no effect. In this case, use {\tt yon}'s IP address in place of its name.) \item Remotely log in to {\tt yon} (for example, via {\tt ssh} or {\tt telnet}), and set the {\tt DISPLAY} environment variable on {\tt yon} to point back to your display. If you use the C-shell on {\tt yon}, this will normally be done by `{\tt setenv DISPLAY hither:0.0}'. Users of {\tt sh}, {\tt ksh}, or {\tt bash} should type `{\tt DISPLAY=hither:0.0; export DISPLAY}'. (Note: if {\tt yon} doesn't recognize {\tt hither} by name, this command won't work. In this case, use {\tt hither}'s IP address in place of its name.) \item Check the connection by starting an X11 client such as {\tt xterm} on {\tt yon}. A new window should appear on your display within a few seconds. (See the documentation for your X server if this doesn't work). \item Once you have succeeded in the test in the previous step, start \WAVE{} in the usual way (`{\tt wave -r} \emph{record ...}') on {\tt yon}. See \hyperref{``Where do I find the missing fonts?''} {``Where do I find the missing fonts?'' (section~} {, page~\pageref{faq:missing-fonts})} {faq:missing-fonts} if your X server complains about missing fonts. See \hyperref{``I can't find the file named \emph{record}!''} {``I can't find the file named \emph{record}!'' (section~} {, page~\pageref{faq:cannot-find-record})} {faq:cannot-find-record} if \WAVE{} complains about missing files. \end{itemize} X servers are standard on all current UNIX workstations, but not on PCs or Macintoshes. There are many MS Windows-hosted X servers available commercially, as well as a smaller number of MS-DOS and Macintosh-hosted X servers. \subsection{Can I run \WAVE{} remotely using a modem?} \index{modem access}\index{PPP}\index{SLIP}\index{TERM} It is possible, with appropriate (SLIP, PPP, or TERM) software, to run \WAVE{} across a serial connection (see the previous question for details), but this will be rather slow even with a 56 kbps modem. Using \WAVE{} remotely via Ethernet is usually quite tolerable, however. If you run \WAVE{} locally, but read data from a remote web site such as PhysioNet, you will have much better performance, whether using a modem or Ethernet, because the volume of data transferred is far lower in this case. \subsection{How can I insert annotations using a two-button mouse?} \index{mouse!middle button!simulating} \index{middle mouse button!simulating} \index{mouse!right button!simulating} \index{right mouse button!simulating} \index{chording (mouse technique)} Normally, the middle button is used to insert annotations, so the problem is to simulate a middle button click if you don't have a middle button. Most, if not all, X servers that support two-button mice provide a way to do this. The most common method is to press both buttons simultaneously (an operation sometimes called ``chording,'' by analogy with playing a chord on the piano). Typically, you have an adjustable window of 50 to 100 milliseconds in which to press both buttons, so that you don't need exactly simultaneous clicks (which would be impossible to achieve reliably). Some X servers delay reporting a button-down event until a button-up event occurs; this approach makes chording less error-prone, but it means that the feedback that \WAVE{} produces in response to button-down events gets delayed until it may no longer be useful. Another approach, more commonly used with one-button mice, is to simulate the middle button (and, if necessary, the right button) using a mouse button and keyboard key combination, such as clicking while pressing the \keycap{SHIFT} or \keycap{CTRL} keys. Check your X server's manuals for information about which of these methods is supported. With a little practice, annotation editing can be tolerable with a button-impaired mouse. Independent of the X server, \WAVE{} also makes it possible to simulate a middle button click, by using the \keycap{F2} key (or the \keycap{5} key on the numeric keypad). If you use this technique, you might also wish to use the \keycap{F3} and \keycap{F4} keys to drag the pointer and marker bars left or right. Inexpensive three-button mice, trackballs, and touchpads manufactured by Logitech and many others are widely available for PCs, and are highly recommended if you do much annotation editing. Most are fully compatible with Microsoft two-button mice. Some users prefer trackballs for precision editing since there is no tendency for the pointer to move when clicking the buttons, as with a mouse. \subsection{How do I get spot help if I don't have a {\sf Help} key?} \index{on-line help}\index{help} \label{faq:no-help-key} If you are not using \WAVE{} with a Sun keyboard, use the \keycap{F1} key to open XView \index{XView} spot help for \WAVE{}. If this doesn't work, you may need to use the `{\tt xmodmap}' utility (a standard component of X11, usually found in the same directory as other X clients such as `{\tt xterm}'). Try the following command: \begin{verbatim} xmodmap -e "keysym F1 = Help" \end{verbatim} If you are now able to use \keycap{F1} to open spot help, you may wish to include this command in your `{\tt .xinitrc}' (a text file in your home directory; create it if it doesn't exist) so that spot help via \keycap{F1} is enabled whenever you log in. If you are not using {\tt olwm} or {\tt olvwm}, \index{olwm@{\tt olwm} (Open Look Window Manager)} \index{olvwm@{\tt olvwm} (Open Look Virtual Window Manager)} spot help may not work properly; in particular, your window manager may pass incorrect information about which window is active when you invoke spot help. There is no general solution to this problem; try using {\tt olvwm} or {\tt olwm}, at least until you are familiar with \WAVE{}'s controls. \section{Problems starting \WAVE{}} \subsection{Why won't \WAVE{} run?} Here are a few things to check: \begin{enumerate} \item \index{wave command@{\tt wave} command} \index{record!name} \index{annotator name} The command used to run \WAVE{} must be typed in lower case, as in `{\tt wave -r 100s -a atr}'. Note that letters in record names and annotator names are also usually in lower case. \item \index{PATH environment variable}\index{environment variables!PATH} Try typing {\tt wave} with no command-line arguments. You should see a summary of options. If you don't, {\tt wave} has not been installed properly, is not in your {\tt PATH}, or another program named {\tt wave} is in your {\tt PATH}. \index{dynamically linked libraries} \index{shared libraries} \index{libraries!missing} \index{LD\_LIBRARY\_PATH environment variable} \index{environment variables!LD\_LIBRARY\_PATH} Errors similar to `{\tt wave: can't load library 'libxview.so.3'}' indicate an installation problem which is probably shared by other applications that use the same libraries. Use the command `{\tt ldd `which wave`} to identify which dynamically linked libraries are missing. Locate these on your system (for example, using a command such as `{\tt find / -name libxview.so.3 -print}'; if any of these libraries cannot be found on your system, it may be downloaded from \htmladdnormallink{PhysioNet}{http://physionet.org/}.) Add the directories in which the missing libraries are found to your {\tt LD\_LIBRARY\_PATH}. For example, if the missing libraries are located in {\tt /usr/openwin/lib}, and you are using the C-shell, use the command `{\tt setenv LD\_LIBRARY\_PATH /usr/openwin/lib:\$\{LD\_LIBRARY\_PATH\}}'. Under Linux, if you can obtain root permissions, simply add {\tt /usr/openwin/lib} to the list of directories in `{\tt /etc/ld.so.conf}', and then run {\tt ldconfig}. This solves the problem permanently (well, at least until the next OS upgrade). \item If the previous test works, try typing `{\tt wave -r 100s -a atr}'. (All \WAVE{} distributions come with record 100s.) One of the following should happen: \begin{itemize} \item \index{WFDB environment variable}\index{environment variables!WFDB} \index{DISPLAY environment variable}\index{environment variables!DISPLAY} If you see a summary of options, rather than the \WAVE{} main window, read the message carefully; you may need to set the {\tt DISPLAY} or {\tt WFDB} environment variables, or the X server may not be running. \item If you see a lengthy error message referring to missing fonts, see \hyperref{``Where do I find the missing fonts?''} {``Where do I find the missing fonts?'' (section~} {, page~\pageref{faq:missing-fonts})} {faq:missing-fonts}. \item If you see neither a summary of options nor the \WAVE{} main window, the {\tt DISPLAY} environment variable may be set incorrectly, the X server may be refusing permission to open a window, or your window manager may be waiting for you to specify the location or size of the window to be opened. \item If you see only a message box reading `{\sf Record 100s is unavailable}', your {\tt WFDB} environment variable may be set incorrectly; in this case, find the directory (probably {\tt /usr/database}) containing a file named `{\tt 100s.hea}' and append the name of that directory to the value of {\tt WFDB}. \item \index{signal!window} If \WAVE{}'s signal window appears, but is solid white or light grey, and you are using a version of \WAVE{} older than 6.8, your X server does not have backing store enabled. Click on any of the navigation controls (e.g., \button{\tt <<}, \button{\tt <}, \button{\tt >}, or \button{\tt >>}), or resize the window, to make the signals appear. To avoid this problem entirely (until the next time you upgrade your X server), turn on backing store in your X server. For instructions for doing this, and for what to do if the signal window is solid (or nearly solid) blue or black, or has only horizontal lines across it, see \hyperref{``I can't see the signals!''} {``I can't see the signals!'' (section~} {, page~\pageref{faq:cannot-see-signals})} {faq:cannot-see-signals}. \index{S option for WAVE@{\tt -S} option for \WAVE{}} \index{command-line options!{\tt -S}}\index{options!{\tt -S}} \item If the signals appear, but there are no annotations or time stamps in the signal window, your X server has defective support for overlay graphics. Restart \WAVE{}, adding the `{\tt -S}' option to the `wave' command. Also see \hyperref{``I can't see text in the signal window!''} {``I can't see text in the signal window!'' (section~} {, page~\pageref{faq:cannot-see-text})} {faq:cannot-see-text}. \item If you see a correct display, your problem is specific to the record you were previously trying to view. Find the directory containing the header file (`\emph{record}{\tt .hea}', where \emph{record} is the record name) for the desired record, and append the name of that directory to the value of {\tt WFDB}. \end{itemize} \end{enumerate} \subsection{Why does \WAVE{} take so long to display the first screen?} If you are running \WAVE{} remotely, a slow network connection may be the reason for sluggish performance. If possible, run \WAVE{} locally, or get a faster network connection. You might also try changing the time scale so that fewer points need to be drawn on each screen. If you are using \WAVE{} to view a record via FTP, it is currently not possible to begin until the entire record has been downloaded; for long records over slow network connections, this can take a long time. Some web (HTTP) servers suffer from the same limitation. If possible, use a web server that supports HTTP range requests, such as Apache, or download a copy of the record to your local disk. PhysioNet and its mirrors support HTTP range requests, so if you experience this problem when reading data from PhysioNet, the cause may be network congestion; in this case, try using a different mirror. \subsection{Where do I find the missing fonts?} \label{faq:missing-fonts} \index{Open Look fonts}\index{font!Open Look} \WAVE{} is an XView \index{XView} application, and requires the Open Look cursor font on the server. If your server doesn't have this font (`{\tt olcursor.snf}' for X11R4 and earlier servers, `{\tt olcursor.pcf}' for X11R5 and later servers), you must install it before \WAVE{} will run. XView applications also use the `{\tt olgl??.snf}' or `{\tt olgl??.pcf}' glyph fonts; if these are missing, \WAVE{} will run but certain standard Open Look graphical elements will not have the correct appearance. To install these fonts on your X server's system: \begin{enumerate} \item Generate copies of `{\tt ol*.snf}' or `{\tt ol*.pcf}' for your X server system's architecture. \item Copy `{\tt ol*.snf}' or `{\tt ol*.pcf}' to a world-readable directory on the server machine. \item Update the font database on the server machine. \item Force the server to reread the font database. \end{enumerate} You should be able to find `{\tt ol*.snf}' or `{\tt ol*.pcf}' files on the machine on which \WAVE{} resides (probably in {\tt /usr/lib/X11/fonts/misc}). The `{\tt ol*.pcf}' files are portable (although usually optimized for the architecture of the system on which they are installed). If you need to use `{\tt snf}' fonts, however, and the server machine is not of the same architecture, it will be necessary to find `{\tt ol*.bdf}' in the XView \index{XView} distribution and use `{\tt bdftosnf}' to generate `{\tt .snf}' files for your X server system's architecture (see the man page for `{\tt bdftosnf}'; this step can be done trivially on the server machine, if `{\tt bdftosnf}' is available there, or with a little more trouble on the remote machine). If the server machine runs UNIX, read the man pages for `{\tt mkfontdir}' and `{\tt xset'} to see how to update and reread the font database; otherwise, consult the documentation for your X server. Note that (under UNIX) you do not need to have write permission in the standard font directory in order to perform these steps; if you add your own font directory to the font path using `{\tt xset}', however, remember that this setting lasts only until the server exits and will need to be repeated afterwards. On some platforms, there are simpler ways to install the missing fonts. See the \htmladdnormallink{XView}{http://physionet.org/physiotools/xview/} home page on PhysioNet for pointers. \section{Display-related questions} \subsection{I can't see the signals!} \label{faq:cannot-see-signals} \index{signal!window} If the signal window is solid white or light grey, and your version of \WAVE{} is older than 6.8, your X server does not have backing store enabled. Click on any of the navigation controls (e.g., \button{\tt <<}, \button{\tt <}, \button{\tt >}, or \button{\tt >>}), or resize the window, to make the signals appear. If this works, the problem is almost certainly that backing store has been disabled in the X server's configuration file. You can verify this by running the command \begin{verbatim} xdpyinfo | grep backing \end{verbatim} To enable backing store, insert the line \begin{verbatim} Option "backingstore" \end{verbatim} in the {\tt Device} section of your X server's configuration file (usually {\tt /etc/X11/xorg.conf} or {\tt /etc/X11/XF86Config}), or run the server with the option ``{\tt +bs}'' to obtain the same result. See the documentation for your X server for further information. \index{format!of signal files} If the signal window is solid or nearly solid blue (or black if you have a mono\-chrome or greyscale display), the signals are too big. \WAVE{} may be attempting to display signals that fill the entire window. In rare cases, noise in the signals themselves can produce this effect. More common causes include incorrect signal format or gain specifications in the header file for the record you have opened, \index{amplitude scale}\index{scales!amplitude} incorrect display calibration, or a choice of amplitude scale that results in excessive vertical range of one or more signals. This effect can also occur as a result of various problems related to the WFDB calibration file, for example, if the {\tt WFDBCAL} environment \index{WFDBCAL environment variable}\index{environment variables!WFDBCAL} variable does not name an accessible WFDB calibration file (see \htmladdnormallink{{\tt wfdbcal(5)}}{../wag/wfdbca-5.htm} for details), if any of the signals in the record are of types not listed in the WFDB calibration file, or if the WFDB calibration file does not contain appropriate default display scales for all of the signal types in the record. See the next several questions for suggestions on correcting these problems. If the signal window contains horizontal lines (in blue if you have a color display) where the signals should appear, the signals are too small. In this case, the signal gains specified in the header file may be incorrect, the display calibration may be incorrect, or you may have specified an amplitude scale that reduces the vertical range of the signals to zero. Click on \button{View...} to check the scales, and adjust them if necessary. If the signal window contains time indicators in the lower corners, look for the signal names along the left edge of the window. If there are no signal names visible, click on \button{View...}, then on {\sf signal names} and \button{Redraw} in the {\sf View} window. If signal names still do not appear, either the header file or the signal file for the record you have opened may be inaccessible. Find these files (see \hyperref{``File-related questions''} {``File-related questions'', section~} {, page~\pageref{faq:file-related-questions}} {faq:file-related-questions}), add the directory that contains them to your WFDB path, and try to read a few samples using `{\tt rdsamp -r \emph{ record} -t 1}'. If this fails, examine the header file for the record (a text file), and be sure that it specifies at least one signal in an accessible signal file. \subsection{I can't see text in the signal window!} \label{faq:cannot-see-text} \index{S option for WAVE@{\tt -S} option for \WAVE{}} \index{command-line options!{\tt -S}}\index{options!{\tt -S}} This may result from an X server bug. This bug appears to be limited to X servers that support `true color' (typically 24- or 32-bit) displays, when using overlay graphics. Restart \WAVE{}, adding the `{\tt -S}' option to the end of the `{\tt wave}' command. \index{Netscape!color map problem with}\index{color map} This problem may also occur on 8-bit `pseudo-color' displays, if another application such as Netscape has already allocated most of the color map. To avoid this problem, start \WAVE{} before starting Netscape. \subsection{Why does \WAVE{} crash, saying ``All pty's in use''?} \index{pty problems} This problem occurs if the installed version of the XView library requires BSD ptys, and if your operating system does not support BSD ptys. \WAVE{} may operate without problems until you attempt to open a text or terminal window (by selecting {\sf Analyze...} from the {\sf File} menu, or by various selections from the {\sf Properties} and {\sf Help menus}. Typically, \WAVE{} then exits with the error ``{\tt All pty's in use}''. Two types of ptys (pseudo-terminal devices) have been supported by most Linux distributions until recently. Traditionally, XView has used BSD-style ptys (pseudo-terminal devices) to implement text windows and terminal emulator objects such as WAVE's analysis commands window. Most recent Linux distributions based on version 2.6 kernels, including Fedora Core 2 and later, support only the newer UNIX98 (SVR4-style) ptys. Although some older XView libraries can be installed on these platforms, they will cause WAVE (and other XView-based applications) to crash whenever a text or terminal window is opened. The recommended solution is to install XView libraries that use UNIX98 (SVR4) ptys rather than the older BSD (``legacy'') ptys; these are available from PhysioNet. \subsection{\WAVE{} doesn't draw/erase properly in the scope window!} \index{Scope window@{\sf Scope} window} This problem may result from several different X server bugs. You may be able to avoid it by switching from overlay graphics mode (the usual default) to shared color mode (selected using \WAVE{}'s {\tt -S} \index{S option for WAVE@{\tt -S} option for \WAVE{}} \index{command-line options!{\tt -S}}\index{options!{\tt -S}} command-line option), or vice versa, since \WAVE{} uses different techniques for drawing in the scope window depending on the display mode. A bug in network file handling can sometimes interfere with correct display of waveforms in the scope window when reading records via HTTP. For now, the only workarounds are to restart \WAVE{}, or to copy the record to local disk files (this can be done easily using \htmladdnormallink{{\tt xform}}{../wag/xform-1.htm}, \htmladdnormallink{{\tt snip}}{../wag/snip-1.htm}, or a web browser). Check \htmladdnormallink{PhysioNet}{http://physionet.org/} for updates to \WAVE{} or to the WFDB library that may correct this bug. \subsection{How can I get correct display scales?} \label{faq:display-scales} \index{xdpyinfo command@{\tt xdpyinfo} command} \index{resolution!of display} Use `{\tt xdpyinfo}' (a standard X client that is provided as part of the X core distribution) to check several server parameters. (As with virtually all X clients, it doesn't matter which machine you find `{\tt xdpyinfo}' on; it will query your server from wherever it is run.) `{\tt xdpyinfo}' will report the version and release numbers of your server; you should have version 11, release 3 or later. It will also report the screen size in pixels and the resolution in pixels per inch. Take a minute to verify the screen resolution by direct measurement of the screen with a ruler. If the resolution reported by `{\tt xdpyinfo}' is incorrect, \WAVE{} will not be able to draw waveforms at properly calibrated scales. The sample X11 servers from MIT and the X Consortium, and the XFree86 project's X11 servers, all support a \index{dpi option for WAVE@{\tt -dpi} option for \WAVE{}} \index{command-line options!{\tt -dpi}}\index{options!{\tt -dpi}} `{\tt -dpi}' option, which allows you to specify the correct screen resolution at the time the server is started; read the man page for X or Xserver for details. (If your server does not support this feature, \WAVE{} can be given a similar `{\tt -dpi}' option.) \index{display calibration}\index{calibration!display} You can also double-check your display calibration using \WAVE{} itself. Run \WAVE{}, enable the grid display, and set the display scales to their default values (25 mm/sec, 10 mm/mV). (Click on \button{View...} to pop up the controls for the grid and the display scales; remember that any adjustments you make become effective only after you click on \button{Redraw} within the {\sf View} window.) The grid intervals should measure exactly 5 mm in each direction. If they are too small, use a larger value for the `{\tt -dpi}' specification; if the grid lines are more than 5 mm apart, decrease the `{\tt -dpi}' values. \subsection{Why are signals too big or too small?} \index{signal!calibration}\index{calibration!signal} There are several reasons why the size of signals may be incorrect. \WAVE{} determines display scales for signals based on several parameters: \begin{itemize} \item \index{WFDBCAL environment variable}\index{environment variables!WFDBCAL} First, the {\tt WFDBCAL} environment variable specifies the name of a text file (located in a directory in the WFDB path) containing the names of many common signals and customary display scales for each (expressed as physical units per centimeter). For example, ECG signals are customarily displayed at a scale of 1 mV per centimeter. If a signal appears at an inappropriately large or small scale, its name may be missing from the calibration file, the {\tt WFDBCAL} variable may not correctly specify the name of the calibration file, or the {\tt WFDBCAL} file may not reside in a directory in the WFDB path. On most systems, the calibration file is `{\tt /usr/database/wfdbcal}', and the {\tt WFDBCAL} environment variable is set by the same command used to set the WFDB path (see the discussion about the \hyperref{database environment} {database environment in section~} {, page~\pageref{sec:wfdb-path}} {sec:wfdb-path}). Instructions for adding additional signal names and scales to the calibration file are located within the file itself, and may also be found in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. \item \index{header file} Second, the header file for the record specifies the gain for each signal (the number of analog-to-digital units per physical unit). If the gain has not been determined, or is incorrect, the size of the signal as drawn by \WAVE{} may be incorrect as well. If you know the physical values that correspond to at least two levels represented in the signal (for example, the top and bottom of a calibration pulse), you can use the `{\tt calsig}' application to determine the gain and to rewrite the header file. See the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm} for details. `{\tt calsig}' can be most conveniently run from \WAVE{}'s {\sf Analysis} window. \item Finally, the display calibration determines the number of pixels per inch (25.4 mm). \WAVE{} usually gets its information about the display calibration from the X server, but the X server may not supply the correct information. See \hyperref{``How can I get correct display scales?''} {``How can I get correct display scales?'' (section~} {, page~\pageref{faq:display-scales})} {faq:display-scales} for details on diagnosing and correcting this problem. \end{itemize} \subsection{How can I enlarge or reduce the size of individual signals?} \WAVE{} version 6.11 and later versions allow you to change the displayed sizes of individual signals if you wish. If the {\sf Annotation Template} is not open, open it by left-clicking anywhere in the signal window, then left-click again in the signal window to return the focus there. Select the signal of interest by pressing \keycap{Shift} and clicking with the left mouse button while the pointer is near the signal of interest, which will be highlighted by \WAVE{} when you do so. Next, press and hold \keycap{Control} and one of \keycap{+} (to enlarge the signal), \keycap{-} (to reduce the size of the signal), or \keycap{=} (to reset the size of the signal to the value selected in the {\sf View} panel). You may repeat this operation with as many signals as desired. If you have not selected any signal, all signals are resized by these commands. \subsection{How can I change \WAVE{}'s default scales or display colors?} \index{scales} The easiest way to make simple changes in \WAVE{}'s default display parameters is by using the controls in the {\sf View} window to configure the display to your liking. If you then click on \button{Save as new defaults} in the {\sf View} window, your choices are recorded in your `{\tt .Xdefaults}' file (in your home directory) and will be used for future \WAVE{} sessions. (Be sure that you have made your changes effective by using \button{Redraw} before \button{Save as new defaults}.) \index{colors!in signal window} Colors are also determined by X resources that may be specified in your `{\tt .Xdefaults}' file, but the {\sf View} window does not include color controls. \hyperref{Read about color resources} {See section~} {, page~\pageref{setting-colors}} {setting-colors} for guidance on setting display colors. \subsection{Why do colors in other windows change when I use \WAVE{}?} \index{S option for WAVE@{\tt -S} option for \WAVE{}} \index{command-line options!{\tt -S}}\index{options!{\tt -S}} \WAVE{} queries the X server to determine how many colors it can display, and what methods it can use to do so. If the X server does not answer this query correctly, \WAVE{} may attempt to use a suboptimal or unavailable display mode. If you find that the colors of other windows change when the cursor moves into a \WAVE{} window, you may prefer to use the `{\tt -S}' option to avoid this behavior at a slight (often unnoticeable) cost in speed. \subsection{Everything in the {\sf Analysis Commands} window appears twice!} \index{stty command@{\tt stty command}} \index{analysis commands window@{\sf Analysis Commands} window} This problem, which may occur when you are running \WAVE{} under Linux, appears to result from a bug in an early Linux version of the XView `termsw' package. Type the command `{\tt stty -echo}' in the {\sf Analysis Commands} window, and the problem should go away. \section{File-related questions} \label{faq:file-related-questions} \subsection{I can't find the file named `\emph{record}'!} \label{faq:cannot-find-record} \index{record!name!vs. file name} \index{file names, vs. record names} Record names are \emph{not} file names; they are components of file names only. For example, three files belong to record 100s: {\tt 100s.hea}, {\tt 100s.dat}, and {\tt 100s.atr}. There is no \emph{file} named `{\tt 100s}' belonging to \emph{record} 100s. See \htmladdnormallink{{\tt annot(5)}}{../wag/annot-5.htm}, \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm}, and \htmladdnormallink{{\tt signal(5)}}{../wag/signal-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for information about file types and formats. \subsection{Why does \WAVE{} tell me `Record ... is unavailable'?} \index{WFDB environment variable}\index{environment variables!WFDB} In common with other WFDB applications, \WAVE{} searches for its input signals and annotations in directories named by the {\tt WFDB} environment variable (see the discussion of the \htmladdnormallink{WFDB path}{../wpg/database-path.htm} in the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm}). If \WAVE{} is running remotely, remember to set {\tt WFDB} \emph{on the \WAVE{} host} \index{WAVE host@\WAVE{} host} (a default can usually be set for C-shell users by `{\tt source /usr/bin/cshsetwfdb}', or for Bourne shell, Korn shell, and `{\tt bash}' users by `{\tt .~setwfdb}'; don't omit the `{\tt .}'). The directories named by {\tt WFDB} are those on the \WAVE{} host; thus (unless your WFDB files are already on the \WAVE{} host, or are accessible to the \WAVE{} host via HTTP or FTP -- see the next question) you must either transfer them to the remote machine, or NFS- or RFS-mount the directory in which they reside onto a suitable point in the file system of the \WAVE{} host. Furthermore, remember that output annotation, log, and print files will be written to your current directory on the \WAVE{} host. \index{wfdbwhich command@{\tt wfdbwhich} command} To get the pathname of a WFDB file, use `{\tt wfdbwhich}' (see \htmladdnormallink{{\tt wfdbwhich(1)}}{../wag/wfdbwh-1.htm} for details). For example, the output of `{\tt wfdbwhich header 100s}' is the pathname of the header file for record 100s (usually `{\tt /usr/database/100s.hea}'). If `{\tt wfdbwhich}' can't find the file, neither can \WAVE{} (or any other WFDB application). In this case, you will need to check and correct the value of the {\tt WFDB} environment variable. In unusual cases, the WFDB path may be correct, but you may not have permission to read the file or one of the directories in the path to the file; in such cases, you will need to get the file's owner or your system administrator to give you appropriate permissions. \subsection{How can I view a record stored on a web site or an FTP server?} Add the directory containing the record to your WFDB path, then use \WAVE{} just as you would to view a record stored on your local disk. For example, suppose the WFDB path is: \begin{verbatim} . /usr/data http://www.signals.r.us/data ftp://sand.bar.com/pub \end{verbatim} In this case, \WAVE{} would look for input files first in the current directory ({\tt .}) of the local disk, then in {\tt /usr/data} (which might be local or might be part of a networked file system), then on the {\tt signals.r.us} web site, and finally on {\tt sand.bar.com}'s FTP server. Unless you have a slow network connection, you may not notice any difference between viewing locally and remotely stored records. \section{Questions about editing} \subsection{Where are my annotations?} \index{annotation file} When you create a new set of annotations, if you have specified the annotator name in the {\sf Load} window, \WAVE{} saves your annotations in a new file with a name of the form {\it record.annotator}, in the current directory. When you edit an existing set of annotations, \WAVE{} makes a copy of the annotation file containing your edits, gives it a name of the form {\it record.annotator}, and saves it in the current directory (i.e., whatever directory was current when you started \WAVE{}). To be able to read your edited copy in a later \WAVE{} session, be sure that the directory that contains your edited copy is listed in your WFDB path \emph{before} any other directory that contains an older version of the same annotation file. Usually, your WFDB path begins with `{\tt .}', a synonym for the current directory. If this is the case, simply return to the directory that contains your edited annotation file before starting \WAVE{} each time. If you created an annotation file without specifying an annotator name, \WAVE{} uses its own name as the annotator name, so you should look for a file named {\it record}{\tt .wave} in this case. Old versions of \WAVE{} created annotation files with names of the form {\it annotator.record}; if you have any of these, rename them in order to continue using them with \WAVE{}. \subsection{Why does \WAVE{} tell me `You may not edit annotations ...'?} Since \WAVE{} is often used to \emph{view} annotated data, without necessarily editing annotations, annotation editing is disabled when \WAVE{} starts. If you have inadvertently done something (such as clicking the middle mouse button in the signal window) that \WAVE{} interprets as an editing command, it will warn you that the editing action is disabled. If you did intend to edit the annotations, select {\sf Allow editing} from the \menubutton{Edit} menu, then repeat the edit. \subsection{How do I edit a record on a CD-ROM, a web site, or an FTP server?} \index{CD-ROM} Be sure that your current directory is writable, that both your current directory and the source (CD-ROM, web, or FTP) directory containing the record you wish to edit are in your WFDB path, and that your current directory comes \emph{before} the source directory in your WFDB path. When \WAVE{} writes your edits, they always go into the current directory (so the current directory cannot be on the CD-ROM, since the CD-ROM is not writable). At a later time, when \WAVE{} (or any other WFDB application) opens the record, it finds your edited version first, and doesn't look for the original version, provided only that, in your WFDB path, the source directory follows the one in which your edits are stored. \subsection{How can I undo an edit?} \index{edit!undo} \WAVE{} doesn't support a general mechanism for `undoing' edits. Once you move an annotation or change its attributes, you must manually restore its position and attributes. Deleted annotations can always be restored, however, by selecting the `phantom' annotation left behind (these are displayed with a `{\sf .}' for the annotation mnemonic) and then `deleting' the phantom. This action restores the attributes of the original annotation, including the {\tt subtype}, {\tt chan}, {\tt num}, and {\tt aux} fields; if the annotation was moved, however, its original position ({\tt time} field) is not restored. \index{annotation file!backup}\index{backup!of annotation file} \WAVE{} is careful not to destroy the version of the annotation file that was current at the time \WAVE{} loaded the record. \WAVE{} always saves edited annotation files to the current directory (even if the original annotation file was loaded from another directory). \WAVE{}'s saved annotation files are given names of the form `{\it record.annotator}'. If, by saving its output, \WAVE{} would overwrite an existing file with the same name, \WAVE{} first renames the existing file by appending a tilde (`{\tt ~}') to its name. Thus it is always possible to recover the state of the annotation file as it was before you began the most recent editing pass. To do so, exit \WAVE{}, and: \begin{itemize} \item If there is a file with a name of the form `{\it record.annotator}\verb+~+' in the current directory, rename it as `{\it record.annotator}'. \item Otherwise, delete or rename `{\it record.annotator}' so that the previous version (which must be in another directory in the WFDB path) becomes visible once again to \WAVE{}. \end{itemize} \subsection{Can I define my own annotations?} \index{anntab file} Yes. See \hyperref{Creating and using new annotation types} {section~} {, page~\pageref{sec:creating-annotation-types}} {sec:creating-annotation-types}, or the file named `{\tt anntab}' in the \WAVE{} distribution for examples. \subsection{How can I recover work after a crash?} \index{autosaving} \WAVE{} autosaves your edits at 1-minute intervals, or after every 20 insertions, deletions, or attribute changes (repositioning doesn't count). You don't need to do anything special to recover your work, but you may lose up to a minute's worth of edits. \subsection{Can I edit annotations with a text editor?} Believe it or not, sometimes users ask this question. \index{wrann command@{\tt wrann} command} \index{rdann command@{\tt rdann} command} \index{annotation!editing using {\tt emacs}} If you have unusual requirements (for example, if you want to change all annotations of a certain type), you may be able to edit annotations more efficiently using a text editor such as {\tt emacs}. Since annotation files are not text files, however, you will need to convert your files from binary to text form to edit them, and then convert the edited text back to binary form in order to use them further with \WAVE{}. You can do this using {\tt rdann} and {\tt wrann} (see \htmladdnormallink{{\tt rdann(1)}}{../wag/rdann-1.htm} and \htmladdnormallink{{\tt wrann(1)}}{../wag/wrann-1.htm}, in the {\it WFDB Applications Guide}, for details). \subsection{What are `link' annotations?} \index{link annotation}\index{annotation!link} By analogy to hyperlinks in World Wide Web documents, link annotations allow external data (including documents, images, and audio annotations among other possibilities) to be associated with specific locations in a record. In fact, any URL that can be understood by a supported web browser can be used to specify the location of external data in a link annotation. Within the link annotation, the URL appears as the first part of the {\tt aux} string (in the {\sf Text} field of the {\sf Annotation Template}). If the {\tt aux} string contains any spaces, any characters following the space are treated as a label to be displayed by \WAVE{} in place of the URL itself. Support for link annotations is a new feature of \WAVE{} 6.0. To view the external data associated with a link annotation, select the link and press \keycap{Enter} (or \keycap{Return}). If your web browser is not already running, this may take a few moments while \WAVE{} starts it. If nothing happens after a minute or so, check \WAVE{}'s {\sf Analysis Commands} window for errors that may have occurred when \WAVE{} attempted to start your web browser. (\WAVE{} uses Mozilla by default. See \hyperref{\WAVE{} and the Web} {section~} {, page~\pageref{sec:web},} {sec:web} for details on configuring \WAVE{} to use a different web browser.) \index{Mozilla}\index{web browser} Link annotations may be moved, copied, attached to specific signals, inserted, changed, and deleted just as for any other annotation. \WAVE{} does not include built-in facilities for editing the external data, however; other software such as a text editor or a special-purpose HTML editor may be useful for this purpose. \subsection{Can \WAVE{} edit signal files?} \WAVE{} itself doesn't do this, but certain kinds of signal file editing are easy to do using {\tt snip} from \WAVE{}'s {\sf Analyze} panel. A common requirement is to remove unwanted data from the beginning or end of a record. To do this, simply mark the segment to be retained using the `{\tt <}' and `{\tt >}' markers, then click on \button{Extract segment}. If your record contains unwanted signals, remove them from the {\sf Signal list} before extracting the segment. You can also rearrange the order of signals within the signal list if you wish. \index{snip command@{\tt snip} command} Note that this operation does not modify the original record; rather, the desired portions of the record are copied to create a new record. If you wish to copy a set of annotations as well as data from signal files, be sure to load the annotation file for the original record into \WAVE{} before extracting the segment. (See \htmladdnormallink{{\tt snip(1)}}{../wag/snip-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, if you need to copy two or more sets of annotations.) \index{wfdbcollate command@{\tt wfdbcollate} command} \index{xform command@{\tt xform} command} To join two or more records end-to-end, use {\tt wfdbcollate} to create a multi-segment record. {\tt wfdbcollate} writes a special header file for the new record, but does not copy or modify the signal or header files of the original records; it will, however, write new annotation files on request, since the WFDB library does not support multi-segment annotation sets. The records to be joined must be ordinary (not multi-segment) records. If necessary, \htmladdnormallink{{\tt xform}}{../wag/xform-1.htm} can be used to rewrite a multi-segment record as an ordinary record. See \htmladdnormallink{{\tt wfdbcollate(1)}}{../wag/wfdbco-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for further information. \index{rdsamp command@{\tt rdsamp} command} \index{wrsamp command@{\tt wrsamp} command} More complex editing of signal files, such as splicing segments or modifying individual samples, can be done by converting the signals to text form using \htmladdnormallink{{\tt rdsamp}}{../wag/rdsamp-1.htm} (accessible from \WAVE{} via the \button{List samples} button in the {\sf Analyze} window), editing the text as required, and converting the edited text to a signal file using {\tt wrsamp} (see \htmladdnormallink{{\tt wrsamp(1)}}{../wag/wrsamp-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details). \section{What else can \WAVE{} do?} \subsection{How can I make a screen dump?} \index{PostScript} \index{xwd command@{\tt xwd} command} \index{xpr command@{\tt xpr} command} \index{screen dump} If you really want a screen dump, `{\tt xwd | xpr}' will produce one in PostScript form. (Type `{\tt man xwd}' and `{\tt man xpr}' for details on options.) Another way to do this, if {\tt xpr} is not available, is `{\tt xwd | xwdtopnm | pnmdepth 255 | pnmtops}'. If your printer is not a PostScript printer, you can still print the output of either of these commands using Ghostscript, a freely available PostScript interpreter. \index{pschart command@{\tt pschart} command} Why settle for a screen dump, though? In most cases, you will prefer to use `{\tt pschart}' to produce a much nicer plot, and in much less time. Select {\sf Print} from \WAVE{}'s \menubutton{File} menu to print the current contents of the signal window; use \button{Print chart} in the {\sf Analyze} window if you wish to specify start and stop times. See \htmladdnormallink{{\tt pschart(1)}}{../wag/pschar-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm} for information about {\tt pschart}'s numerous formatting options. If you don't like the defaults, change them by editing \WAVE{}'s menu file. Doing so also allows you to change the default format for the {\sf Print} choice on the \menubutton{File} menu; see the comments in the menu file for details). \index{psfd command@{\tt psfd} command} If you are plotting a great deal of data, you may wish to use \htmladdnormallink{{\tt psfd}}{../wag/psfd-1.htm} rather than {\tt pschart}; to do so, select \button{Print full disclosure} rather than \button{Print chart} in \WAVE{}'s {\sf Analyze} window. Most of {\tt pschart}'s options are accepted by {\tt psfd}. If you need to collect a set of figures, either from a single record or from many records, use \WAVE{}'s {\sf Log} window to record the times of interest, and enter captions for each figure in the {\sf Description} field of the {\sf Log} window. The log file generated in this way can be interpreted directly as a command file by {\tt pschart}, which prints the captions as titles for each figure. \subsection{Can \WAVE{} read digitized signals in \emph{(fill in the blank)} format?} \index{format!of signal files}\index{file formats!signals} \index{signal file!formats} Probably it can. The most common formats for storage of digitized signals are fixed-length binary formats that encode integer samples, and \WAVE{} supports dozens of such formats. \WAVE{} does not require any embedded header information within signal files (but it can be made to ignore such information). In most cases, all you need to do is to write a header file that describes the format in terms understandable to \WAVE{}; once you have done so, any of the other WFDB applications can also read your signals. Writing a header file can be most easily done by copying an existing header file and modifying it using a text editor. See \htmladdnormallink{{\tt signal(5)}}{../wag/signal-5.htm} and \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm} in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details on supported signal formats and how to write a header file. \index{multiplexed signal file} \index{block-interleaved signal file} There are at least three common ways of storing multiple signals: in separate files, in multiplexed (sample-interleaved) files, and in block-interleaved files. In a multiplexed file containing N signals, each set of N consecutive samples contains one sample from each signal. In a block-interleaved file containing N signals in blocks of 512 samples, for example, the first 512 samples come from the first signal, the next 512 samples come from the second signal, and so on. \WAVE{} supports reading multiple signals from separate files or from multiplexed files (or both in the same recording), but it does not directly support block-interleaved files in general. \WAVE{} does support reading a special type of file containing signals sampled at different sampling frequencies. Files of this type contain ``frames'' of samples. Each frame contains at least one sample of each signal, but there may be two or more samples of signals sampled at multiples of the frame rate within each frame. Block-interleaved files may be described simply as files with very large frames. If you have a block-interleaved file, therefore, you may be able to view it with \WAVE{} via this expedient: \begin{itemize} \item Write a header file describing the layout of the block-interleaved signal file (see \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details). The sampling frequency in the first line of the header should be given as the frame rate (i.e., the basic sampling rate divided by the number of samples per block), and the ``samples per frame'' field in each signal specification line should be given as the number of samples per block. \index{resolution!of multi-frequency records} \index{high-resolution mode} \item \index{WFDB\_HIGHRES@{\tt WFDB\_HIGHRES}} Open the record in ``high-resolution'' mode with \WAVE{} (see \hyperref{``What is high-resolution mode?''} {section~} {, page~\pageref{faq:high-res}} {faq:high-res}). \end{itemize} \index{resolution!of time in annotation files} \index{format!of signal files!changing} One significant limitation of this approach is that the time resolution of any annotation files created in this way is limited to the frame interval. You can avoid this limitation by reformatting the signal file (for example, by following the steps above to verify that the signal file is properly readable, then using {\tt xform} with its \index{xform command@{\tt xform} command} own {\tt -H} option; see \htmladdnormallink{{\tt xform(1)}}{../wag/xform-1.htm} in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. \index{embedded header data in signal files} \index{preamble in signal files} \index{signal file!preamble in} A common feature of many signal file formats is \emph{embedded header data}: bytes at the beginning of the signal file that do not contain digitized samples. By specifying the length of the embedded header data as the byte offset for the sample data in the (external) header file, such formats can be easily accommodated; \WAVE{} and any other applications built using the WFDB library simply ignore any preamble. If the signal file format is documented, it is usually not difficult to write a program to generate a header file based on embedded header data. See \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details on byte offsets. \WAVE{} does not support directly reading signals stored in text files, because there is no general method for computing the position within such files of samples occuring at arbitrarily-chosen times. It is usually simple to convert such text files into \WAVE{}-compatible binary files using \htmladdnormallink{{\tt wrsamp}}{../wag/wrsamp-1.htm}, \index{wrsamp command@{\tt wrsamp} command} an application supplied with \WAVE{}. For the same reason, \WAVE{} does not support directly reading signals stored using variable-length encoding or non-uniform sampling intervals. If a decompression utility is available, convert the data to fixed-length, uniform samples. \WAVE{} does not support any non-integer storage formats either, mainly because such formats are not commonly used for digitized data. If you have signals in an unsupported format, you have two choices if you wish to analyze them using \WAVE{}: \begin{itemize} \index{format!of signal files!changing} \item Write a utility to convert your data to a supported format. This is usually easy to do -- use `{\tt wrsamp.c}' (the source for {\tt wrsamp}, provided with \WAVE{}) as a model. \item Add support for your format to the WFDB library, and recompile the shared WFDB library. Once the new library is installed, \WAVE{} and the other WFDB applications will be able to use your format without recompilation. If you have a large amount of data, this may be the only reasonable choice. It may not be much more difficult than writing a conversion utility, provided that your format permits computing file position as a function of time. See the appendix titled `{\sf Extensions}' in the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm} for hints on how to proceed. \end{itemize} \index{EDF (European Data Format)} For example, \WAVE{} does not support directly reading EDF, the European Data Format first used for recordings of polysomnograms. (This is \emph{not} the format used for either the European ST-T Database or the MIT-BIH Polysomnographic Database, both of which can be read directly by \WAVE{}.) EDF uses block interleaving, and also includes an embedded header. Although the method outlined above for reading block-interleaved files can be used to read EDF files, a difficulty in doing so is that the block size is specified within the embedded header, and may vary between records. Moreover, since typical block sizes are quite large, the time resolution of annotations in this case is very coarse. A format converter for EDF files, \htmladdnormallink{{\tt edf2mit}}{http://physionet.org/}, may be used to rewrite such files in a format that \WAVE{} can read without these limitations. This converter also constructs a suitable header file from the embedded header in the EDF file. All currently supported signal formats use 16 bits per sample or less. In principle, formats requiring up to 32 bits per sample should pose no problem in environments that support \WAVE{}, but formats requiring more than 16 bits may be difficult to support in other environments. \subsection{What is ``high-resolution'' mode?} \label{faq:high-res} \index{high-resolution mode} \index{resolution!of multi-frequency records} Briefly, it is an alternative display mode for multi-frequency records only. See the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm} for details on multi-frequency records. \index{H option for WAVE@{\tt -H} option for \WAVE{}} \index{command-line options!{\tt -H}}\index{options!{\tt -H}} The {\tt -H} option, introduced in \WAVE{} 6.0, selects high-resolution mode. For ordinary records (those for which all signals are sampled at the same frequency), high-resolution mode is identical to standard mode. \index{resolution!of display} \index{oversampled signal}\index{signal!oversampled} In a multi-frequency record, some signals are sampled at multiples of the basic sampling rate. These are referred to as \emph{oversampled signals}. In standard mode, \WAVE{} \emph{decimates} these signals (i.e., \WAVE{} reduces their sampling rates to the basic sampling rate by averaging successive samples) before displaying them. In high-resolution mode, however, \WAVE{} displays each sample from any oversampled signals if the screen resolution is high enough (note that at standard display scales, it may be difficult to see the difference). In this mode, \WAVE{} replicates additional copies of each sample of any signals sampled at lower rates before displaying them; in consequence, particularly at highly magnified display scales, these signals may have a marked stairstep appearance. \index{resolution!of time in annotation files} Note that the time resolution for placement of annotations is the same in high-resolution mode and in standard mode. Thus, for example, it is not possible to place annotations on consecutive samples of an oversampled signal, since the second annotation will replace the first one. This is a limitation of the WFDB library, not of \WAVE{} itself, and may be removed in a future release of the WFDB library. \subsection{Can \WAVE{} open more than one record at once?} \index{signal!window} \index{r option for WAVE@{\tt -r} option for \WAVE{}} \index{command-line options!{\tt -r}}\index{options!{\tt -r}} Yes. All records should be listed in one command-line argument, following the {\tt -r} argument on the command line; separate the record names by `{\tt +}' characters. For example, the command \begin{verbatim} wave -r 237+237n & \end{verbatim} starts a \WAVE{} \emph{process group}, opening a separate \WAVE{} signal window for each of record {\tt 237} and record {\tt 237n}. Each signal window corresponds to a separate \WAVE{} process; thus you may browse through and edit each record independently. The signal window assigned to the last record in the group (in this case, {\tt 237n}) is the \emph{master signal window}; it contains a \button{Sync} button, which may be used to synchronize the other signal windows with the master at any time. By synchronizing, we mean that the other windows are redrawn as necessary so that the times shown in their lower left corners match that shown in the master signal window. It is also possible to synchronize by clicking on a point of interest in the master signal window. To do this, press and hold \keycap{Ctrl} while clicking with the middle mouse button. The master signal window will be unaffected, but the other windows are redrawn so that they begin at the time corresponding to the point of interest. This feature is particularly useful if you have a derived record such as in the heart rate signal analysis example (see section~\ref{sec:analysis-example}) open in the master signal window, and the raw signals in another window; you can then click on a feature in the derived signal and instantly see the corresponding raw signals. Remote-control applications such as {\tt wavescript} and {\tt wave-remote} can also start or control a \WAVE{} process group. All signal windows in the group move synchronously in response to commands from {\tt wavescript} or {\tt wave-remote}. You may change the record that is open in any signal window, either via the {\sf Load} window or using a remote-control application. You may quit from any signal window without affecting the others. Note that there are a few important limitations: \begin{itemize} \item The number of signal windows (\WAVE{} processes) in any process group cannot be \emph{increased} once the process group has been created. \item Although you may run two or more \WAVE{} process groups simultaneously, applications such as {\tt wavescript} can control only a single process group at a time (by default, the last one started). \item Although you may have the same record open in more than one signal window, you should not edit the same annotator of the record in more than one window at a time (just as you should not edit the same text file in more than one editor window at once). You may edit different annotators of the same record freely. \end{itemize} Taken together, the first two of these points imply that if you decide after opening a record that you would like to add another record to the group, it will be necessary to exit and start again, naming all desired records in the group on the \WAVE{} command line when you do so. If you wish to move through two records sampled at the same sampling frequency in lockstep (for example, to view digitally filtered signals and the original unfiltered signals side-by-side), another approach is to create a single header file that names the signal files for both records. It is not necessary for all signals to be in the same signal file, or even on the same disk drive. \subsection{Can \WAVE{} open more than 32 signals in a window at once?} Yes. There is no fixed limit on the number of signals that current versions of \WAVE{} can open. Earlier versions were limited to 32 or fewer signals. If the signal window gets to be too crowded, select a subset of signals to be displayed by editing the signal list (in \WAVE{}'s {\sf Analyze} window, and choose {\sf Draw: listed signals only} in the {\sf View} window. \subsection{Can \WAVE{} open more than one annotation file in a window at once?} No. In most cases, however, it's not really necessary to do so. \index{bxb command@{\tt bxb} command} For example, if you want to compare two sets of annotations of the same data in order to study their differences, use {\tt bxb} with the `{\tt -o}' option to prepare a comparison annotation file, and then use \WAVE{} to open that file. (See \htmladdnormallink{{\tt bxb(1)}}{../wag/bxb-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details.) If your goal is to find discrepancies, this is by far the best way to do so; `{\tt bxb}' produces a statistical summary of them, and you can search for {\sf NOTE} annotations to find each disagreement (enter `{\tt "}' in the \index{search}\index{annotation!searching for} `{\sf Search for}' field of \WAVE{}'s {\sf Find} window). If you want to annotate multiple signals independently (for example, ECG and respiration), choose `{\sf attached to signals}' from the \amenubutton{Show annotations:} menu in \WAVE{}'s {\sf View} window. See \hyperref{Multi-edit mode} {section~} {, page~\pageref{sec:multi-edit},} {sec:multi-edit} for details. \index{mrgann command@{\tt mrgann} command} If you already have multiple independent annotation files for a single record (whether these contain annotations for independent signals or annotations for disjoint segments of the record), you can use {\tt mrgann} to merge them into a single file, preserving information about which file each annotation came from if you wish (see \htmladdnormallink{{\tt mrgann}}{../wag/mrgann-1.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, for details). \index{wfdbcollate command@{\tt wfdbcollate} command} \htmladdnormallink{{\tt wfdbcollate}}{../wag/wfdbco-1.htm} can also be used for this purpose if you have separate annotation files for each segment of a multi-segment record. \subsection{Can \WAVE{} do smooth scrolling?} No. \subsection{Can \WAVE{} be used for real-time display of data being acquired?} Not easily. If you have a UNIX application that can write data in a format readable by \WAVE{}, it should be possible in principle to run it from \WAVE{}'s {\sf Analyze} window, and to use \WAVE{} to review the data being acquired, using the \button{Reload} button in \WAVE{}'s {\sf Analyze} window to read additional samples as they become available. (\WAVE{} will not attempt to read beyond what it thinks is the end of the record; if it ever catches up with the digitizing process, therefore, \WAVE{} can only be forced to read further by using \button{Reload}, or by closing and reopening the record.) \subsection{Can \WAVE{} scroll through a record without user intervention?} \label{sec:autoscroll} Yes. Create a log file that contains entries spaced at the desired intervals throughout the record. A suitable interval might be 10 seconds if that is the width of the signal window. Open the {\sf Log} window, load the log file, and click on \button{\tt >>} to start the scrolling. Adjust the delay between frames using the slider on the {\sf Log} window. \index{Scope window@{\sf Scope} window} Another way to review a record is via the scope window (accessible by clicking on \button{Show scope window} in \WAVE{}'s {\sf Analyze} window). Note that the scope display, since it is triggered by annotations, is of no use in detecting erroneously unannotated waveforms (false negatives), since such waveforms are simply not displayed in the scope window. The scope is useful for studying morphologic variation of waveforms, and for locating erroneously annotated waveforms (false positives). \subsection{How can I use \WAVE{}'s menu variables in a script or other program?} See \WAVE{}'s default menu file for examples of passing \WAVE{}'s menu variables as command-line arguments to other programs. \index{menu variables} Click on \button{Export variables} in \WAVE{}'s {\sf Analyze} window if you need to use these variables within a shell script. (If you are not using {\tt bash}, {\tt ksh}, or {\tt sh}, edit the \WAVE{} menu file first to select the alternate definition of this action.) Once the variables have been exported, they may be used in the same way as any other environment variables within your shell scripts. For example, the time of the left edge of the signal window is {\tt \$LEFT}. Within a C program, read the values of these variables using {\tt getenv}; for example, {\tt getenv("LEFT")} returns the (string) value of the time of the left edge of the signal window. Note that the exported values are only a snapshot of the variables at the time you click on \button{Export variables}; the values do not track later actions performed within \WAVE{}. \subsection{Is there a Motif version of \WAVE{}?} No. (Rant begins here.) When I designed \WAVE{}, I attempted to use Motif, but switched to XView when it became clear that much of the documented functionality of the Motif toolkit (at the time, version 0.9) was unimplemented or unusable. It also seemed unlikely that anyone would prefer Motif's buggy, ugly, proprietary, bloated, kludgy procedural interface over XView's buggy, attractive, non-proprietary, streamlined, elegant object-oriented interface (end of rant). \index{Open Look} Motif and its lookalikes (notably GTK+, Qt, and LessTif) have clearly become the standard, however, and a consequence of this is that the Open Look user interface presented by \WAVE{} is unfamiliar to many users. Until recently, there has been a scarcity of introductory material for Open Look in print (but O'Reilly's Open Books Project (\htmladdnormallink{http://www\-.oreilly\-.com/open\-book/open\-look/} {http://www.oreilly.com/openbook/openlook/}) now provides free on-line copies of three relevant books (a user's guide and two programmer's guides). Unfortunately, the Motif API, and those of its lookalikes, are vastly different from the XView API with which \WAVE{} is written, and a port to Motif would be decidedly non-trivial. \subsection{How can I find out about ...?} There are several sources of information about \WAVE{}: \begin{itemize} \index{PostScript} \item If you have not used \WAVE{} before, start by reading the \WAVE{} ~{\it User's Guide} (this book) and working through the tutorial examples it contains. If you need to use existing WFDB applications under \WAVE{}'s control, refer to the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. If you need to develop your own applications for use with \WAVE{}, read the \htmladdnormallink{{\it WFDB Programmer's Guide}}{../wpg/wpg.htm}. You may read these guides on-line on-line or print your own copies (visit \htmladdnormallink{PhysioNet}{http://physionet.org/}). \index{on-line help}\index{help}\index{Mozilla}\index{web browser} \item The on-line version of the \WAVE{} ~{\it User's Guide} can be opened for browsing by clicking left on \button{User's Guide} in \WAVE{}'s {\sf Help Topics} window (this requires \WAVE{} 6.0 or later, and Mozilla (or another browser, if named by the {\tt URLV} environment variable) on the \WAVE{} host). This guide may also be read with any Web browser independently of \WAVE{} (point your browser to \htmladdnormallink{{\tt file:///usr/help/html/wug/wug.htm}} {file:///usr/help/html/wug/wug.htm} if you have installed \WAVE{} 6.0 or later on your system, or to \htmladdnormallink {{\tt http://physio\-net.\-org\-/physiotools\-/wug/}} {http://physionet.org/physiotools/wug/} otherwise). \item Use XView \index{XView} spot help for brief descriptions of \WAVE{}'s controls and other graphic elements. Point to any control and press the \keycap{Help} key (see \hyperref{``How do I get spot help ...?''} {``How do I get spot help ...?'', section~} {, page~\pageref{faq:no-help-key},} {faq:no-help-key} if you don't have a \keycap{Help} key). \item Use \WAVE{}'s task-oriented on-line help to learn how to use \WAVE{} as a tool for interactive data analysis, control of external programs, annotation editing, and other topics. Click left on \button{Help} in \WAVE{}'s main control panel to open the {\sf Help} window, containing buttons for each topic. Click left on any of these buttons to open a help window for the associated topic. Within the help window, use the scroll bar to move through the topic, or click on \button{Print} button to obtain a printed copy of the entire topic. \item Select `{\sf About WAVE...}' from the \menubutton{Properties} menu, or \button{What's new} from the {\sf Help} window, for a summary of new features in \WAVE{}, a current list of known bugs, and instructions for printing a copy of the on-line manual, which includes the text of all of the other help topics. Click on \button{Print} to obtain a printed copy of the `{\sf What's new}' topic. \item The {\tt man} page for \WAVE{} (\htmladdnormallink{{\tt wave(1)}}{../wag/wave-1.htm}) contains a concise description of its command-line options, the environment variables and X resources it uses, and the \WAVE{} menu file. (This information is also included in this Guide; the {\tt man} page is intended as a compact reference. The {\tt man} page for \WAVE{} can be found in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}, or on-line by typing `{\tt man wave}', or by using {\tt xman} or {\tt tkman}, etc.) \item For additional information on the \WAVE{} menu file, see the comments in the default menu file distributed with \WAVE{} and usually installed as `{\tt /usr/lib/wavemenu.def}'. \item For information about the data formats supported by \WAVE{} and the other WFDB applications, see \htmladdnormallink{{\tt annot(5)}}{../wag/annot-5.htm}, \htmladdnormallink{{\tt header(5)}}{../wag/header-5.htm}, \htmladdnormallink{{\tt signal(5)}}{../wag/signal-5.htm}, and \htmladdnormallink{{\tt wfdbcal(5)}}{../wag/wfdbca-5.htm}, in the \htmladdnormallink{{\it WFDB Applications Guide}}{../wag/wag.htm}. \end{itemize} \chapter{Command-line Options} \index{starting WAVE@starting \WAVE{}}\index{wave command@{\tt wave} command} To start \WAVE{}, open a terminal window and type into it: {\tt wave -r} \textit{record options \ldots} \noindent In most cases, \textit{record} is a single record name; you may, however, name several records to be opened by concatenating their names with {\tt +} between them, as in `{tt wave -r 100+101+102 -a atr}', which opens records {\tt 100}, {\tt 101}, and {\tt 102}, each in its own signal window. Note that there must be no spaces around the `+' characters (the list of records should appear as a single command-line argument without embedded spaces). If the {\tt -r} \textit{record} arguments are omitted, \WAVE{} simply prints a summary of its command-line syntax in the terminal window. The \textit{options} may include any of these, in any order: \begin{description} \item[{\tt -a} {\it annotator}] Open the specified annotation file for \textit{record}. If this file does not exist, it will be created once you insert an annotation. \index{resolution!of display} \item[{\tt -dpi} {\it xx}\textrm{{[}{\tt x}}{\it yy}\textrm{]}] Calibrate \WAVE{} for use with a display having a resolution of \textit{xx} (by \textit{yy}) dots per inch. \item[{\tt -f} {\it time}] Open the record beginning at the specified \textit{time}. \item[{\tt -g}] Use shades of grey only, even on a color monitor. \index{resolution!of multi-frequency records} \item[{\tt -H}] Operate in ``high-resolution'' mode when reading multi-frequency records. \item[{\tt -m}] Use monochrome (usually black and white) only, even on a color or greyscale monitor. The line styles selected by the {\tt -m} option may be easier to distinguish on some greyscale monitors than the default shades of grey. \item[{\tt -O}] Use overlay graphics for maximum speed and display quality if possible. This is the usual default if the X server supports a PseudoColor or GrayScale visual. This option exists only to force use of overlay graphics if a different mode has been chosen as the default. \item[{\tt -s} {\it signal} \textrm{[} {\it signal \ldots} \textrm{]}] Initialize the signal list. By default, the signal list includes all available signals, in numerical order. \item[{\tt -S}] Use the standard (shared) color map, even if is possible to modify the color map. Using this option conserves color resources if you have other applications that use non-standard colors, at a small (possibly unnoticeable) expense of some speed in redrawing the display. The {\tt -S} option may be used in combination with the {\tt -g} option if desired. \end{description} The following group of options can be used to enable the optional display elements normallly chosen in the {\sf View} window. \begin{description} \item[{\tt -Vs}] Display annotation {\tt subtyp} fields. \item[{\tt -Vc}] Display annotation {\tt chan} fields. \item[{\tt -Vn}] Display annotation {\tt num} fields. \item[{\tt -Va}] Display annotation {\tt aux} fields. (This is the default for annotations such as rhythm changes and comments; selecting this option causes {\tt aux} fields to be displayed for all annotations, and also causes the annotation mnemonics to be displayed for rhythm changes, comments, etc.) \item[{\tt -Vm}] Display annotation marker bars. \item[{\tt -VN}] Display signal names along the left edge of the signal window. \item[{\tt -Vb}] Display baselines for any DC-coupled signals, and label the zero levels and the units along the right edge of the signal window. \item[{\tt -Vl}] While the pointer is in the signal window and any mouse button is depressed, track the intersections of the marker bar with the signals and draw horizontal marker bars across the signal window at the levels of these intersections. \end{description} The final group of options allows you to select the initial choices from the menu buttons in the {\sf View} window. Each of these options should be followed by a numeric argument that matches the \emph{position} of the desired choice in the menu, where the top item on each menu is in position 0, the one below it is in position 1, etc. \index{amplitude scale}\index{scales!amplitude} For example, to set the initial amplitude scale to 5 mm/mV (the item at position 2 in the {\sf Amplitude scale} menu), add {\tt -Vv 2} to the command line. \begin{description} \index{time scale}\index{scales!time} \item[{\tt -Vt} {\it n}] Set the time scale (0: 0.25 mm/hour, 1: 1 mm/hour, 2: 5 mm/hour, 3: 0.25 mm/min, 4: 1 mm/min, 5: 5 mm/min, 6: 25 mm/min, 7: 50 mm/min, 8: 125 mm/min, 9: 250 mm/min, 10: 500 mm/min, 11: 12.5 mm/sec, 12: 25 mm/sec (default), 13: 50 mm/sec, 14: 125 mm/sec, 15: 250 mm/sec, 16: 500 mm/sec, 17: 1000 mm/sec, 18: 2000 mm/sec, 19: 5000 mm/sec, 20: 10 mm/ms, 21: 20 mm/ms, 22: 50 mm/ms, 23: 100 mm/ms, 24: 200 mm/ms, 25: 500 mm/ms). \index{amplitude scale}\index{scales!amplitude} \item[{\tt -Vv} {\it n}] Set the amplitude scale (0: 1 mm/mV, 1: 2.5 mm/mV, 2: 5 mm/mV, 3: 10 mm/mV (default), 4: 20 mm/mV, 5: 40 mm/mV, 6: 100 mm/mV). \item[{\tt -VS} {\it n}] Set the choice on the {\sf Draw} menu (0: all signals (default), 1: listed signals only). \item[{\tt -VA} {\it n}] Set the choice on the {\sf Show annotations} menu (0: centered (default), 1: attached to signals, 2: as a signal). \item[{\tt -VT} {\it n}] \index{absolute time} \index{elapsed time} \index{time!display mode} Set the choice on the {\sf Time display} menu (0: elapsed (default), 1: absolute, 2: in sample intervals). \item[{\tt -VG} {\it n}] Set the choice on the {\sf Grid} menu (0: none, 1: 0.2 s, 2: 0.5 mV, 3: 0.2 s x 0.5 mV, 4: 0.04 s x 0.1 mV). \end{description} \WAVE{} also accepts generic XView options, including the {\tt -default} and {\tt -xrm} options for setting \hyperref{X11 resources} {X11 resources (see section~} {, page~\pageref{sec:resources})} {sec:resources}. \section{Running two or more \WAVE{} processes} By specifying two or more record names, separated by `{\tt +}' characters, in the command-line argument that follows `{\tt -r}' (see above), you may open separate \WAVE{} signal windows (processes) for each record. These processes are almost completely independent: from any signal window, you may navigate within the record, change display settings, edit annotations, run external analysis programs, quit the process, etc., without affecting any other signal windows. For example, you may open two signal windows for the same record by: \begin{verbatim} wave -r 100+100 -a atr \end{verbatim} You can now move about the record freely in either window. This facility makes it easy to compare different segments of the record. Note that whenever two or more windows are displaying the same set of annotations, as in this case, only one should be editing the annotations at any given time. The window associated with the \emph{last} record named on the command line has a special status: it is designated the master signal window, and a \button{Sync} button appears at the top of this window. Clicking on \button{Sync} causes all of the other signal windows to be redrawn so that the times shown in their lower left corners match that in the master signal window. (Note, however, that if you have quit a signal window from the middle of the list, any signal windows from earlier in the list will no longer respond to sync requests.) By default, all command-line arguments apply to all signal windows. You may specify an argument that is to apply to only one signal window, however, by prefixing the argument with `{\tt +}{\it n}{\tt /}', where {\it n} is the \emph{signal window number}. (The first signal window, corresponding to the first record named on the command line, is signal window number 0; the next is number 1, etc.) This facility has many applications. For example, you may wish to open two copies of the same record, with two different annotators: \begin{verbatim} wave -r 100+100 -a +0/atr +1/qrs \end{verbatim} In this case, record {\tt 100} is opened in two windows, with annotator {\tt atr} in window 0 and annotator {\tt qrs} in window 1. (The `{\tt -a}' option applies to both windows since it does not have a `{\tt +}{\it n}{\tt /}' prefix.) As another example, you may wish to discuss a record with colleagues at other locations: \begin{verbatim} wave -r 200+200+200 -a qrs +0/-display +0/atlantic.bigu.edu:0 \ +1/-display +1/pacific.widget.com:0 \end{verbatim} Here, record {\tt 200} is opened in three windows. Window 0 is opened on display 0 of {\tt atlantic.bigu.edu}, window 1 on display 0 of {\tt pacific.widget.com}, and window 3 (the master window) on the local display. (For this to work, your colleagues must first allow your computer to open windows on their displays, typically using {\tt xhost}. See \htmladdnormallink{{\tt xview(7)}}{../wag/xview-7.htm} for information about the {\tt -display} option. Notice that the `{\tt +}{\it n}{\tt /}' prefix must be attached to both the {\tt -display} option and to its argument in order to apply both of these arguments to the same signal window.) Your colleagues can freely move about the record, but you can direct the discussion at any time by using the \button{Sync} button in your signal window. In a case such as this one, anyone can enable editing; you should do so only after making sure that no one else has. Once you have saved your work (by selecting {\sf Save} from the {\sf File} menu), your changes become visible to your colleagues if they reload the annotations (by clicking on {\sf Reload} from the {\sf Load} window). As a final example, the MIMIC Database includes both high-resolution waveform records and medium-resolution (roughly 1 sample per second) computed measurement records. You may view both of these at the same time using a command such as: \begin{verbatim} wave -r 237+237n -a all \end{verbatim} Typically, you will wish to view the high-resolution and low-resolution data at different time scales. Although \WAVE{} attempts to choose reasonable defaults, you can adjust the scales independently if you wish: \begin{verbatim} wave -r 237+237n -a all +1/-Vt +1/2 \end{verbatim} If you use {\tt wavescript} or {\tt wave-remote} to control the master signal window (this happens by default unless you use the {\tt -pid} option of these programs to control a different signal window), the other signal windows are kept synchronized with the master window. Note that you cannot \emph{increase} the number of signal windows in a group once you have started a \WAVE{} process group, although you can run more than one process group at a time if you wish. \chapter{Menu Variables} \label{ch:menu-variables} In \WAVE{}'s menu file, the variables listed below are interpreted by \WAVE{} before they are passed to the shell, typically as command-line arguments in button-action commands. \index{record!name} \index{annotator name} \index{time scale}\index{scales!time} \index{amplitude scale}\index{scales!amplitude} \vspace{.5cm} \begin{tabular*}{\textwidth}{l p{3.5 in}} {\tt \$RECORD} & { the current record name } \\ {\tt \$ANNOTATOR} & { the current input annotator name } \\ {\tt \$START} & { the current {\sf Start (elapsed)} time as shown in the {\sf Analyze} window (the time of the `{\tt <}' marker, if any) } \\ {\tt \$END} & { the current {\sf End (elapsed)} time as shown in the {\sf Analyze} window (the time of the `{\tt >}' marker, if any) } \\ {\tt \$DURATION} & { the time interval between {\tt \$END} and {\tt \$START} } \\ {\tt \$LEFT} & { the time of the left edge of the signal window } \\ {\tt \$RIGHT} & { the time of the right edge of the signal window } \\ {\tt \$WIDTH} & { the time interval between {\tt \$RIGHT} and {\tt \$LEFT} } \\ {\tt \$SIGNAL} & { the number of the selected signal } \\ {\tt \$SIGNALS} & { the signal list } \\ {\tt \$LOG} & { the name of the current log file } \\ {\tt \$WFDB} & { the WFDB path (from the {\sf Load} window) } \\ {\tt \$WFDBCAL} & { the WFDB calibration file (from the {\sf Load} window) } \\ {\tt \$TSCALE} & { the time scale, in mm/sec } \\ {\tt \$VSCALE} & { the amplitude scale, in mm/mV } \\ {\tt \$DISPMODE} & { the annotation display mode (0: annotations displayed in center, no marker bars; 1: annotations displayed in center, long marker bars; 2: annotations attached to signals, no bars; 3: annotations attached to signals, short bars; 4: annotations displayed as a signal, no bars; 5: annotations displayed as a signal, long bars) } \\ \end{tabular*} \begin{tabular*}{\textwidth}{l p{3.5 in}} {\tt \$PSPRINT} & { the command for printing PostScript data from the standard input, as specified in the {\sf Print setup} window } \\ {\tt \$TEXTPRINT} & { the command for printing text from the standard input, as specified in the {\sf Print setup} window } \\ {\tt \$URL} & { the URL specified by the most recently selected link } \end{tabular*} \vspace{.5cm} Other strings that begin with `{\tt \$}' are passed to the shell unchanged. \chapter{Keyboard Command and Mouse Action Summary} \section{General rules for all Open Look applications} \index{Open Look} The basic rules for using a mouse in an Open Look application such as \WAVE{} are: \begin{itemize} \index{mouse!right button!uses for} \item \emph{To open menus (always identified by a $\nabla$ at the right end of the control), click right.} \index{mouse!left button!uses for} \item \emph{To use any other control, click left.} \end{itemize} Notice boxes and some other windows have \emph{default buttons}, identified by a double oval outline. If the pointer is anywhere in the window, you can press \keycap{Enter} (\keycap{Return}) instead of clicking left on the default button if you wish. To type into a text field, move the mouse pointer near the insertion point and click left. The text cursor should appear nearby; if it is a black triangle, you may begin typing. Use \keycap{BackSpace} or \keycap{Delete} to erase the character to the left of the text cursor. To move the cursor within the field use the arrow keys (\keycap{$\leftarrow$}, \keycap{$\rightarrow$}), or use {\tt emacs}-style commands: \begin{description} \item[\keycap{Ctrl}+\keycap{A}] move to the beginning \item[\keycap{Ctrl}+\keycap{E}] move to the end \item[\keycap{Ctrl}+\keycap{F}] move forward one character \item[\keycap{Ctrl}+\keycap{B}] move backward one character \end{description} Until you press \keycap{Return} (or \keycap{Enter}), your typing is not interpreted. Unless spot help is disabled, you can get help on any control by pointing to it and pressing \keycap{Help} (or \keycap{F1} if you don't have a \keycap{Help} key). To enable spot help, see \hyperref{``How do I get spot help ...?''} {``How do I get spot help ...?'', section~} {, page~\pageref{faq:no-help-key},} {faq:no-help-key}. \section{Using the keyboard and mouse in the signal window} In this section, wherever two keyboard actions are specified, the first should work on any keyboard, but the second may be usable on Sun type 4 keyboards only. Mouse actions in the signal window are generally related to selecting and editing annotations. If annotation editing is disabled, actions that would otherwise modify the annotation file are disallowed. \subsection*{Simple mouse actions} These actions are used to select, insert, change, and move annotations. \begin{description} \index{mouse!left button!uses for} \item[left mouse button click] If the {\sf Annotation Template} is visible or iconified, select the annotation to the left of the pointer (if this annotation is not in the signal window, recenter the signal window on it). Otherwise, display the {\sf Annotation Template}. \index{mouse!middle button!uses for} \item[middle mouse button click] Insert an annotation at the location indicated by the pointer, with fields as specified in the {\sf Annotation Template}. If the pointer is within the selection rectangle, the selected annotation is replaced by the inserted one; in this case, if the {\sf Type} is `{\tt .}' (deleted annotation), the selected annotation is marked for deletion. An annotation so marked can be restored by deleting it again. \index{mouse!right button!uses for} \item[right mouse button click] Select the annotation to the right of the pointer (if this annotation is not in the signal window, recenter the signal window on it). \end{description} If the pointer is moved during any of the simple mouse actions above, the selected annotation (if any) is \emph{dragged} along with the pointer. When the button is released, the selected annotation is \emph{dropped} at the new location. \index{marker!editing} Even if editing is disabled, you may insert, delete, or move `{\tt :}', `{\tt <}', and `{\tt >}' markers using these mouse actions (or their keyboard equivalents, below) if you first set the {\sf Type} field of the {\sf Annotation Template} to the desired marker type. If you prefer not to have the {\sf Annotation Template} visible while editing, you can iconify it so that it will not continually pop up whenever you click left. \subsection*{Keyboard equivalents for simple mouse actions} \index{mouse!keyboard equivalents} These actions can be used on any system. If you have a one- or two-button mouse, your X server should provide other methods for simulating middle and right button clicks; see the documentation for your X server for details. \begin{description} \item[\keycap{$\leftarrow$}] Simulate a left mouse button click. \item[\keycap{F2} (or \keycap{5} on the numeric keypad)] Simulate a middle mouse button click. \item[\keycap{$\rightarrow$}] Simulate a right mouse button click. \end{description} To move the pointer left or right, use \keycap{F3} and \keycap{F4} (or \keycap{=} and \keycap{*} on the numeric keypad). To simulate a dragging action, press and hold \keycap{$\leftarrow$}, \keycap{F2}, or \keycap{$\rightarrow$}, use \keycap{F3} or \keycap{F4} to move the pointer to the desired location, then drop the selected annotation by releasing all keys. \subsection*{Mouse and keyboard combination actions} These actions are used to select signals. To perform these actions, press and hold the designated key, perform the mouse action, then release the key. \index{signal!selecting} \begin{description} \item[\keycap{Shift}+left mouse button] Select the signal nearest the pointer (the selected signal is highlighted, and its signal number is entered into the {\sf Signal} field in the {\sf Analyze} window). This action can also be performed using \keycap{Shift}+\keycap{$\leftarrow$}. \item[\keycap{Ctrl}+left mouse button] Select the signal nearest the pointer as above, and also enter it into the {\sf Signal list} field in the {\sf Analyze} window. This action can also be performed using \keycap{Shift}+\keycap{$\leftarrow$}. \index{Meta key@{\sf Meta} key} \item[\keycap{Meta}+left mouse button] Select the signal nearest the pointer as above, but \emph{delete} it from the {\sf Signal list}. (There is no keyboard-only equivalent.) \end{description} (Most keyboards do not have a key labelled {\sf Meta}. On most Sun keyboards, the {\sf Meta} keys look like this: \keycap{$\Diamond$}; they flank the space bar. On most other keyboards, the \keycap{Alt} key is equivalent to {\sf Meta}.) \subsection*{Other keyboard actions} These actions are used for annotation editing and navigation through the record. \begin{description} \item[\keycap{$\uparrow$}] (In multi-edit mode only, and only if an annotation is attached.) Move the selected annotation to the next emph{lower} signal number. If another annotation is at that location, however, select that annotation instead. If (within the {\sf View} window) you have chosen to draw all signals, this corresponds to moving the selection up one signal on the display. If you have chosen to draw listed signals only, the order of signals in the signal list determines where the annotation moves on the display. \item[\keycap{Ctrl}+\keycap{$\uparrow$}] As for \keycap{$\uparrow$} alone, except that the selected annotation is copied rather than moved. \item[\keycap{$\downarrow$}] As for \keycap{$\uparrow$}, except that the selected annotation is moved to the next \emph{higher} signal number (i.e., down one signal on the display if all signals are drawn). \item[\keycap{Ctrl}+\keycap{$\downarrow$}] As for \keycap{$\downarrow$} alone, except that the selected annotation is copied rather than moved. \item[\keycap{F6} (or \keycap{Copy})] Copy the selected annotation into the {\sf Annotation Template}. \item[\keycap{F9} (or \keycap{Find})] \index{search}\index{annotation!searching for} Search forward (as for \button{Search {\tt >}}). \item[\keycap{Ctrl}+\keycap{F9} (or \keycap{Ctrl}+\keycap{Find})] Search backward (as for \button{{\tt <} Search}). \item[\keycap{Shift}+\keycap{F9} (or \keycap{End})] Move to the end of the record. \item[\keycap{Ctrl}+\keycap{Shift}+\keycap{F9} (or \keycap{Home})] Move to the beginning of the record. \item[\keycap{F10} (or \keycap{PgDn})] Move forward half a screen (as for \button{\tt >}). \item[\keycap{Shift}+\keycap{F10} (or \keycap{PgUp})] Move backward half a screen (as for \button{\tt <}). \item[\keycap{Ctrl}+\keycap{F10} (or \keycap{Ctrl}+\keycap{PgDn})] Move forward a full screen (as for \button{\tt >>}). \item[\keycap{Ctrl}+\keycap{Shift}+\keycap{F10} (or \keycap{Ctrl}+\keycap{PgUp})] Move backward a full screen (as for \button{\tt <<}). \item[\keycap{Enter} (or \keycap{Return})] \index{link annotation}\index{annotation!link} (Only if a link annotation has been selected.) Show the external data specified by the link using a web browser; start the web browser first if necessary. \end{description} Commands in the next group are used to change the display scales. If a signal is selected, amplitude changes apply to the selected signal {\em only}. If shown, the grid reflects the scale of signal 0 (the top signal). \index{zoom} \index{enlarging signals} \index{signal!enlarging} \index{reducing size of signals} \index{changing scales} \index{scales!changing} \index{rescaling} \index{time!compressing and expanding} \begin{description} \item[\keycap{Ctrl}+\keycap{(}] zoom in (show more detail by expanding the time axis) \item[\keycap{Ctrl}+\keycap{)}] zoom out (show more context by compressing the time axis) \item[\keycap{Ctrl}+\keycap{+}] increase the amplitude of the signal(s) \item[\keycap{Ctrl}+\keycap{-}] decrease the amplitude of the signal(s) \index{inverting signals}\index{signal!inverting} \item[\keycap{Ctrl}+\keycap{*}] invert the signal(s) \item[\keycap{Ctrl}+\keycap{=}] restore the scales selected on the {\sf View} panel \end{description} Typing an annotation or marker mnemonic sets the {\sf Type} field of the {\sf Annotation Template} (even if the {\sf Annotation Template} is not visible, and even if editing is disabled). These mnemonics are shown on the \hyperref{{\sf Type} menu} {{\sf Type} menu (see figure~} {, page~\pageref{fig:type-menu})} {fig:type-menu}. \newpage \addcontentsline{toc}{chapter}{Index} \printindex \end{document}