mcplot

A Python package with a plotting class and routines for publication-ready graphics.

About mcplot

mcplot provides a class that combines methods to easily produce publication-ready graphics on light or black background. It includes a large number of colormaps collected from different sources. There are a number of functions that help to position plots, number plot panels, or generally write text on a graph.

The complete documentation of mcplot is available at:

https://mcuntz.github.io/mcplot/

Calling a plotting script from the command line

mcplot provides a class that has methods for opening and closing different plotting backends, setting layout options, as well as having a command line interface. A most basic example is using the method plot_test, which just plots to sinusoidal curves. A file mcplot_test.py could be:

# file: mcplot_test.py
from mcplot import mcPlot

if __name__ == '__main__':
    iplot = mcPlot(desc='Test mcPlot',
                   argstr='No argument wanted')
    iplot.plot_test()
    iplot.close()

This script mcplot_test.py can be called on the command line. '-h' gives a short help:

python mcplot_test.py -h

gives the help message:

usage: mcplot_test.py [-h] [-o plotname] [-s] [-t outtype] [-u]
                      [-w] [--dpi number] [--transparent] [args ...]

Test mcPlot

positional arguments:
  args                  No argument wanted

options:
  -h, --help            show this help message and exit
  -o plot_filename, --output plot_filename,
  -p plot_filename, --plotname plot_filename
                        Name of plot output file for types pdf, html,
                        d3, or hvplot, and name basis for type png
                        (default: class_mcplot).
  -s, --serif           Use serif font; default sans serif.
  -t outtype, --type outtype
                        Output type is pdf, png, html, d3, or hvplot
                        (default: open screen windows).
  -u, --usetex          Use LaTeX to render text in pdf, png and html.
  -w, --white           White lines on transparent or black background;
                        default: black lines on transparent or
                        white background.
  --dpi number          Dots Per inch (DPI) for non-vector output types or
                        rasterized maps in vector output (default: 300).
  --transparent         Transparent figure background
                        (default: black or white).

python mcplot_test.py

opens a standard Matplotlib plotting window with the test plot.

python mcplot_test.py -t pdf -p test1.pdf

writes the plot into the PDF file test1.pdf using the sans-serif font DejaVuSans that comes with Matplotlib. It will use the serif font DejaVueSerif with the command line option -s. It will use LaTeX to render text with the -u option. -u -s uses LaTeX standard Computer Modern font. It uses MyriadPro as sans-serif font in LaTeX, which must be installed (see section Myriad Pro).

By default, mcPlot plots onto a DIN A4 page, which facilitates choices of font sizes, etc. The output can easily be cropped with the utility pdfcrop which can be acquired from CTAN. The standard subplots are on a 2x3 grid. The plot will be tightly cropped if the output type is png. Plot resolution can be set for png as well (--dpi) with standard 300 dpi. PNG plots can also have transparent background (--transparent), for example for use in presentations.

The command line switch -w swaps foreground and backgroud colours, i.e. plots white lines on black background. This is used if you do presentations with black background.

In summary, the standard command line options allow to use the same script to design a plot using plotting windows on screen, produce the publication ready plot writing into s PDF file (-t, -p, -u options), and make the same plot with dark background for presentations (-t, -p, -u, -w options).

Using the plotting class

The class mcPlot can be extended. One normally would have at least a method to read data from a file and a method that produces a plot. This could give a script such as:

# file: mcplot_basic.py
import numpy as np
from mcplot import mcPlot


class myPlot(mcPlot):

    def read_data(self):
        # reading one file would use self.cargs[0] such as
        # self.dat = np.loadtxt(self.cargs[0])
        self.dat = np.arange(100)

    def plot_fig_1(self):
        import matplotlib.pyplot as plt

        # make axes
        fig = plt.figure()
        ax = fig.add_subplot(3, 2, 1)

        # plot
        xx = self.dat / float(self.dat.size) * 4. * np.pi
        line1 = ax.plot(xx, np.sin(xx))
        plt.setp(line1, linestyle='-', linewidth=self.lw,
                 marker='', color=self.lcol1)

        # show plot or write in file
        self.plot_save(fig)


if __name__ == '__main__':
    # open plot
    iplot = myPlot(desc='A basic plot')
    # read data
    iplot.read_data()
    # plot
    iplot.plot_fig_1()
    # close plot and possible output file
    iplot.close()

The script could be called giving the name(s) of (an) input file(s) on the command line, which is then accessible through self.cargs:

python mcplot_basic.py -t png -p basic_ input.csv

Every time self.plot_save(fig) is called, a figure is written to the output file. A PDF file can have multiple pages. For PNG files, only the start of the output files is given and will be extended by f'{start}{self.ifig:04d}.png'. The example would give the outputfile basic_0001.png.

See the complete documentation of mcplot at: https://mcuntz.github.io/mcplot/

Installation

The easiest way to install is via pip:

pip install mcplot

or via conda:

conda install -c conda-forge mcplot

Requirements

• numpy
• matplotlib
• pandas

License

mcplot is distributed under the MIT License. See the LICENSE file for details.

Copyright (c) 2021- Matthias Cuntz