Installation¶

$ sudo apt-get install python-rpy2
$ conda install -c r r-essentials
$ conda install -c r r-rjson
$ pip install rpy2

Load the magic¶

In [1]:

%load_ext rpy2.ipython

Create a dataframe¶

Use Python and pandas to create a dataframe.

In [2]:

import pandas as pd
df_from_python = pd.DataFrame({'A': [4, 3, 5, 2, 1, 7, 7, 5, 9],
                               'B': [0, 4, 3, 6, 7, 10,11, 9, 13],
                               'C': [1, 2, 3, 1, 2, 3, 1, 2, 3]})
df_from_python

Out[2]:

	A	B	C
0	4	0	1
1	3	4	2
2	5	3	3
3	2	6	1
4	1	7	2
5	7	10	3
6	7	11	1
7	5	9	2
8	9	13	3

Create plot in R¶

Import the dataframe df with the -i argument.

In [3]:

%%R -i df_from_python
require(ggplot2)
# Plot the DataFrame df
ggplot(data=df_from_python) + geom_point(aes(x=A, y=B, color=C)) + ggtitle('R scatter!')

/home/jitsejan/anaconda3/lib/python3.6/site-packages/rpy2/rinterface/__init__.py:186: RRuntimeWarning: Loading required package: ggplot2

  warnings.warn(x, RRuntimeWarning)

Create a dataframe using R¶

Create a dataframe and export it using -o.

In [4]:

%%R -o df_from_r
d <- c(2, 6, 6, 8, 4, 7, 5, 9, 3)
e <- c(3, 1, 4, 3, 6, 7, 10,11, 9)
f <- c(3, 1, 2, 3, 1, 2, 3, 1, 2)

df_from_r <- data.frame(d, e, f)

Now the variable is available for Python.

In [5]:

df_from_r

Out[5]:

	d	e	f
1	2.0	3.0	3.0
2	6.0	1.0	1.0
3	6.0	4.0	2.0
4	8.0	3.0	3.0
5	4.0	6.0	1.0
6	7.0	7.0	2.0
7	5.0	10.0	3.0
8	9.0	11.0	1.0
9	3.0	9.0	2.0

Create plot with Python¶

In [6]:

import matplotlib.pyplot as plt
import matplotlib
%matplotlib inline
matplotlib.style.use('ggplot')

plt.scatter(x=df_from_r['d'], y=df_from_r['e'], c=df_from_r['f'])
plt.title('Python scatter!')
plt.xlabel('A')
plt.ylabel('B')
fig = matplotlib.pyplot.gcf()
fig.set_size_inches(10, 10)