The PySciPlotTK (Python Scientific Plotting Toolkit) provides a workflow for the creation of plots for papers, theses, posters and presentation. Naturally, it uses the congenial matplotlib.
The features are:
- There are style presets available from which you can choose (e.g. Latex and Matlab). All system settings (font size, line width etc.) are already set so that the plots look really nice.
- Presets for plot sizes are available (e.g. Revtex, A0 poster, PowerPoint presentation, standard Latex document). The plot width is restricted to one (=normal size) or two (=wide size) text columns, because most journals (Physical Review Letters, Physical Review B, Applied Physics Letters etc.) are restrictive concerning plot widths. Also, it looks better anyway if you don't have 50 different plot sizes in your thesis. We ensure that the plot size and font sizes are always consistent.
- It is very easy to replot many plots in a different style, size or file format, e.g. if you want to recreate diagrams for a different journal or a presentation.
- You can supply style/size/file format parameters in the plot script or as command line arguments.
- It works well together with Makefiles, so you can integrate it into your preexisting Latex Makefile workflow.
Requirements: Matplotlib needs to be installed on your system, e.g. with
sudo apt-get install python-matplotlib
Package download and installation:
git clone https://github.com/zonksoft/PySciPlotTK
cd PySciPlotTK
sudo python setup.py install
Package download and installation in user directory:
If you don't have root rights, use:
git clone https://github.com/zonksoft/PySciPlotTK
cd PySciPlotTK
python setup.py install --user
Package download and setting PYTHONPATH:
Use the package in-place by setting $PYTHONPATH (do it only if you know why you want this - e.g.
if you plan to make your own changes and you don't want to run setup.py all the time):
git clone https://github.com/zonksoft/PySciPlotTK
cd PySciPlotTK
echo "export PYTHONPATH=$(pwd):\$PYTHONPATH" >> ~/.bashrc && . ~/.bashrc
- Styles: matlab (sans-serif fonts), latex (serif fonts)
- Sizes: revtex, a0poster, presentation, latexa4
- File formats: whatever matplotlib supports
The main functionality is contained in the EasyPlotter class. You need to supply the output filename and the style. The style has the syntax 'style,size', e.g. 'matlab,revtex', 'latex,revtex', 'matlab,a0poster', 'latex,presentation' etc.
Example 1: Plot a one-column ('normal') figure as PDF for a revtex document in Matlab style.
from pysciplottk.easyplotter import EasyPlotter
plotter = EasyPlotter('output.pdf','latex,revtex')
fig = plotter.normal_figure()
ax = fig.add_subplot(111)
ax.plot([1,2,3],[4,5,6])
plotter.save()Example 2: Plot a two-column ('wide') figure as PNG for a presentation in Latex style.
from pysciplottk.easyplotter import EasyPlotter
plotter = EasyPlotter('output.png','matlab,presentation')
fig = plotter.wide_figure(height=3)
ax = fig.add_subplot(111)
ax.plot([1,2,3],[4,5,6])
plotter.save()Suppose you have a Python script whose only job is to create a plot. You can easily change file format, style and size by supplying different command line arguments.
Plot script plot.py:
import sys
from pysciplottk.easyplotter import EasyPlotter
plotter = EasyPlotter.from_argv(sys.argv)
ax = plotter.normal_figure_single_ax()
ax.plot([1,2,3],[4,5,6])
if plotter.flag == 'plot_more':
ax.plot([4],[19])
plotter.save()By invoking the script with various command line arguments, you can change the result:
python plot.py
python plot.py output.pdf
python plot.py output.pdf latex,revtex plot_more
python plot.py output.png matlab,presentation do_not_plot_more
The most convenient way is to put a lot of plot scripts which use the EasyPlotter.from_argv function into a src directory and use a Makefile to replot changed files or to replot everything with a different style. You can either write your own Makefile that uses the EasyPlotter command line interface or use the one in the tools directory that we wrote for you.
- Copy the supplied Makefile to a directory.
- Copy/create all plot scripts in a subdirectory src and make sure that they all use the EasyPlotter.from_argv constructor.
- Run the Makefile.
Example make commands:
make
make pdf
make eps
make pdf STYLE=latex,revtex
make png STYLE=matlab,a0poster
make eps STYLE=matlab,revtex FLAG=plot_more
All of those commands execute all new/changed plot scripts in the src directory, so that you end up with an up-to-date set of output files.
Every combination of file format, size, style and flag will be created in a different subdirectory, e.g. the last line will create the directory eps-matlab,revtex-plot_more.
Pro tip: A trick to mask the actual style from the perspective of a Latex document is to create a symlink:
ln -s eps-matlab,revtex-plot_more eps
In the Latex document, you can access the files through eps/file.eps. If you decide later that you e.g. like the latex style better than the matlab style and you want to add a flag, you can change the symlink with
ln -s eps-matlab,revtex-do_not_plot_more eps
and just rerun the Latex compiler without having to change the document.