PSRCHIVE user documentation: pav

1.0 Purpose

The Pulsar Archive Viewer, pav, is designed to visually display the data stored in an archive. It uses the PGPLOT graphics subroutine library and any of its associated output devices. Like vap, pav is designed to extract and display information only, it will not modify archives in any way. It does, however, allow for memory-only pre-processing of an archive. This means that any processing options specified on the command line are performed on the archive in memory after it has been read from disk and are lost once the program exits. To permanently modify an archive on disk, use pam. The program is built around the PSRCHIVE Plotter class, which contains a vast and bewildering array of visualisation algorithms. The user can select which type of plot they desire, specify pre-processing and display fine-tuning options, then show the result on screen or direct it to a file according to the drivers installed in the local version of pgplot.

2.0 Usage

pav is fairly simple to use, though the wide range of options available to the user can cause confusion. All commands are of the form:
vap options filenames
where the options can be anything from extensive list of available choices. For a full list of options, use:
pav -h
Multiple filenames can be specified on the command line either individually or using wildcard characters. Each file will be displayed in sequence, one after the other. Alternatively, the view panel can be partitioned into a grid of smaller panels (using the "-N" option) so as to fit multiple files in the same window.

3.0 Algorithms

The PSRCHIVE Plotter class contains a wide range of visualisation algorithms. The simplest of these allows single profiles to be extracted and plotted, more complex algorithms combine multiple dimensions to form a colour or contour map, or a composite of various visualisation styles.

4.0 Testing and examples

pav works by setting a flag for each of the available plot styles and processing options. It will run and exit cleanly even if no visualisation flags are selected, simply producing no output. Visualisation flags are (usually) independent of the pre-processing flags, which allow the user to, say, combine a certain number of phase bins together in order to clean up a noisy profile. Multiple visualisation flags can be combined, usually producing each display in turn. The following examples show various ways in which pav can be used to visualise archives: To combine all the data in an archive into a single integrated pulse profile, use:
pav -DFTp filename
In this case, "D" is the visualisation flag and "FTp" are all pre- processing flags. To view an integrated profile with polarisation information, use:
pav -SFT filename
where the visualisation option "D" has been exchanged for "S". Some of the plotting options are only relevant to certain file types. For example, a plot designed to show instrumental phase will only yield useful information if given a calibrator file. The user must exercise a certain amount of judgement in these cases. It is also possible to fine-tune the graphical output. For example, to produce a plot suitable for publication, it might be necessary to exclude the filename from the heading and change the colour scheme to better suit a light background. The umbrella option "--publn" handles the formatting of labels and headers; there are multiple options for manipulating colour schemes, depending on the type of plot.

5.0 Known bugs and features that require implementation

  • None at this time.