Skip to main content

Matplotlib with Rmarkdown

Ch. 3 Sharing data objects

Last update: Thu Nov 19 17:20:43 2020 -0600 (49b93b1)

One of the advantages of running R and Python code chunks in the same document is that we can share object and variables between the environment of both programming languages. There are functions at what R excels, and functions that run better at Python. We take the best of both worlds.

We can share object from R in Python, or share Python objects in R.

3.1 From R to Python, back to R

  1. Always load the Python environment with the packages you need.

  1. Load the dataset in R and assign it to an R object. Let’s call it autos:

R

# R chunk
autos = cars       # assign cars to autos

  1. Read the R object from Python by adding the prefix r. before the name of the R object autos. Then, assign it to a Python object that we will name autos_py.

Python

# Python chunk
import numpy
import pandas 

autos_py = r.autos    # assign to a Python object

  1. Make a calculation between two columns in the dataset (distance and speed), and assign it to a new column in the dataset with autos_py['time'].

Python

# Python chunk
autos_py['time'] = autos_py['dist'] / autos_py['speed']   # calculate on variables

  1. From R, read the Python object py$autos_py and plot the dataset with the new column time, that you obtained in Python. Observe that we added the prefix py$ in front of the Python object autos_py:

R

# R chunk
plot(py$autos_py)          # plot a Python data object

3.2 From Python to R, back to Python