Using Python's Matplotlib with Rmarkdown

One way to include Matplotlib plots in Rmarkdown files within Blogdown is to save them to the /static directory. Files in this directory are then copied over to the /public directory for publishing.

pwd
mkdir -p ../../static/post/2017-10-28-using-python-s-matplotlib-with-rmarkdown_files
# The p flag will create the directory (and intermediate subdirectories) if it doesn't exist, otherwise will do nothing

## /Users/zac/git/zacbook.net/content/post

knitr::current_input() # will return the name of the Rmd file

## [1] "2017-10-28-using-python-s-matplotlib-with-rmarkdown.Rmd"

getwd() # will return the current work directory

## [1] "/Users/zac/git/zacbook.net/content/post"

rprojroot::find_rstudio_root_file() # will return the root of the project

## [1] "/Users/zac/git/zacbook.net"

# So should ensure that a directory is created in 
# (ROOT)/(current work directory, removing project root, replace 'content' with 'static')/filename/
# Could also, perhaps, simply replace 'content' with 'static' in current work directory?
# although it seems that blogdown does not make subdirectories within 'static' to put images from rcode, so perhaps need to simplify to projectroot/static/filename ?
# The above statement is incorrect - blogdown does create subdirectories within static. The version that was not in a subdirectory was an old version!

For example:

import numpy as np
import matplotlib.pyplot as plt
n = 256
X = np.linspace(-np.pi, np.pi, n, endpoint=True)
Y = np.sin(2*X)
fig, ax = plt.subplots(nrows=1, ncols=1)
ax.plot (X, Y+1, color='blue', alpha=1.00)

## [<matplotlib.lines.Line2D object at 0x14f9e2f40>]

ax.plot (X, Y-1, color='blue', alpha=1.00)

## [<matplotlib.lines.Line2D object at 0x14f9e2a00>]

fig.savefig(
    (
        '../../static/post/'
        '2017-10-28-using-python-s-matplotlib-with-rmarkdown_files/foo.png'
    ),
    bbox_inches='tight',
    dpi=600
)

Then using

![A wavy plot.](/post/2017-10-28-using-python-s-matplotlib-with-rmarkdown_files/foo.png)

the plot will be included in the post. Note that the subdirectory 2017-10-28-using-python-s-matplotlib-with-rmarkdown needed to be created by hand. Knitr will not create these automatically as is done with plots generated from R code.

Output image:

A wavy plot.

This method does have limitations, such as breaking the inline previews in Rstudio, since they are expected to be in the same directory as the *.Rmd file.

Using Python's Matplotlib with Rmarkdown

Zac Harmany