quick & dirty plot

qdp is a shell script that interfaces with pyxplot so as to generate scientific plots from the commandline. As pyxplot itself, qdp has some defaults that should satisfy the user more often than not. If it is not the case, there are plenty of command-line options that control how the plots are generated. The default output consists of both PDF and PNG files with the resulting plot.

In any case, as its name suggest, qdp is quick and dirty so if something does not work as you expect, feel free to modify it to suit your needs. It was originally written to enhance the scriptability of wasora.

Installation

Copy or symlink the file qdp to a binary path, such as /usr/local/bin or $HOME/bin (provided it is in the PATH environment variable). Make sure that the pyxplot executable can be found system-wide:

$ which pyxplot
/usr/bin/pyxplot

If the output of which pyxplot is empty, please first install pyxplot.

Usage

Check the usage and available options with the -h option:

$ qdp -h
usage: ./qdp [options] [-o output] [datafile]
 no datafile indicates stdin
 if no output is given, basename is assumed equal to datafile
  if neither output nor datafile given, basename output is qdp
  either .png and .pdf will be appended to output
     --ycol "<ints>"         set y columns in datafile, default all
     --xcol <int>            set x column in datafile, default 1
     --xlabel <string>       set x axis label, default none
     --ylabel <string>       set y axis label, default none
     --xrange <range>        set x axis range, default auto
     --yrange <range>        set y axis range, default auto
     --grid                  set grid, default no
     --width <float>         set plot width in cm, default 12
     --with "<strings>"      set plotting styles, default linespoints
     --lt "<strings>"        set linetypes, default 1
     --lw "<floats>"         set linewdiths, default 1
     --plw "<floats>"        set pointlinewidths, default 1
     --ps "<float>"          set pointsizes, default 1
     --pt "<ints>"           set pointtype, default some values
     --pi "<ints>"           set pointinterval (extension to pyxplot)
     --color "<strings>"     set colors, default blue, black, green, etc.
     --ti "<strings>"        set column titles, default none
     --plottitle <string>    set plot title, default none
     --key <string>          set key location, default above
     --pre <string>          execute pyxplot commands in  <string> before plotting
     --pdf                   generate only pdf output, default both
     --png                   generate only png output, default both
     --dpi <float>           dpi to use in png output, default 127
     --header <file>         use <file> as template instead of default
     --debug                 leave the .ppl file for inspection and edition

 quotes are needed if several columns are to be plotted
 see the README.md file and the examples subdirectory

A simple call to qdp with datafile as an argument will plot all the columns of datafile versus the first one with lines adding a bullet at a random interval. The result will be written to two files named datafile.pdf and datafile.png with dots in datafile replaced to dashes to avoid having two extensions (LaTeX complains when calling \includegraphics otherwise). If the first line of the datafile contains a commented (with a hash #) line with the column titles, they are automatically used by qdp.

The subdirectory examples contains some data files that illustrate how qdp can quickly build aesthetically-pleasant plots. For example, go to the examples subdirectory and execute from the shell

$ qdp lag.dat

If everything went fine, two figures should have been created: lag-dat.pdf and lag-dat.png, which are ready to be included into your LaTeX article or into your web page (respectively).

Note that the data file lag.dat was generated using the tool wasora from the input file lag.was:

$ wasora lag.was > lag.dat

The script qdp can be fed through a pipe directly from wasora as

$ wasora lag.was | qdp

In this case, as no -o option was given, the outputs are qdp.pdf and qdp.png.

The default behavior can be modified using commandline arguments. For example,

$ qdp lag.dat -o biglag --png --dpi 600 --color "orange olivegreen" --lw "3 1" --ti "signal\$(t)\$ lag\$(t)\$" --key top

Note the quoting and the escaping of the arguments to the options of qdp: each argument should consist of a single shell token, hence to tell qdp which colors are to be used, the names orange and olivegreen should be quoted to form a single token (otherwise only the orange would be understood as the color to be used for column two, and olivegreen would be taken as an input data file). In the same sense, the dollar sign is a special character for the shell so it should be escape in order to actually pass a dollar sign to qdp (which in turn passes it to LaTeX).

Further examples

The following invocations of qdp use the example data files contained in the examples subdirectory.

The Lorenz attractor vs time:

$ qdp lorenz.dat --color "brickred navyblue gray" --pi "60 71 87"

Numerical integrals of algebraic expressions:

$ qdp algebraic.dat --ti "\$f(x)\$ \$g(x)\$ \$\\int_0^x{f(x')g(x')\,dx'}\$"

Numbers of primes less than a certain number:

$ qdp nprimes.dat --ti "real \$n/\\ln(n)\$ \$\\int_2^n\\frac{dt}{\\ln(t)}\$" --xlabel "\$n\$" --key "bottom" --pi "1 1 1"

Distribution of thermal-hydraulic properties of an arbitrarily-powered vertical boiling channel:

$ qdp arbitrary-steady.dat --ti "\$q^\star(z)\$ \$u^\star(z)\$ \$h^\star(z)\$ \$\rho^\star(z)\$" --lw "2 2 1 1" --key bottom --pre "set axis x atzero" --xlabel "\$z\$" --yrange "[-1:2.1]"

qdp is copyright (c) 2013-2014 jeremy theler jeremy@talador.com.ar
licensed under GNU GPL version 3 or later.
wasora is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.