如何直接互相使用python与R

作者：赵相杰，中科院遗传与发育生物学研究所生物信息学研究生，xjzhao@genetics.ac.cn

很多时候不同的问题用各自适合的工具去解决会更舒服（比如一次性的数据处理工作交给shell，数据处理、机器学习交给Python，统计和可视化交给R）。但是不同的语言有各自专用的IDE，比如python的Pycharm、Spyder等，R的Rstudio，在不同的IDE之间来回切换比较容易让人分心。即使有jupyter notebook中irkernel，可以允许在jupyter notebook中创建以R为kernel的notebook，不同notebook间导出、导入仍然让我们很不爽。我们需要一个可以在一个文件中写不同程序的方法。这篇文章介绍了如何利用ipython的魔法方法在Python和R之间轻松切换

Warning: 个人认为该方法适合日使用频率 Python/R > 7/3的做数据分析的同学。比如日常工作是使用python做数据导入导出、数据处理和建模等，用R做统计和可视化，甚至部分统计和可视化也是由Python完成

Requirements:

conda: a package and environment manager

ipython: for interactive computing (using ipython magic method)

rpy2: a library for accessing R within Python

1. Installation

In temina

>conda create -n test python=3.7 ipython rpy2 simplegeneric pandas  # 安装rpy2的适合会顺便把R装到该环境下，环境的位置在anaconda_dir/envs/（如果conda是anaconda自带的话）>conda activate test>ipython  # 或者安装这个环境的jupyter notebook/jupyter lab (个人推荐)，见references中的ipykernel

# 在jupyter notebook/lab中或者ipython的交互式界面（jupyter notebook是加强版ipython，jupyter lab是jupyter notebook的加强版/未来版）# load ipython's rpy2 extension before calling rmagic methods%load_ext rpy2.ipython# %load_ext是ipython的魔法方法之一，调用后就可以用%%R或者%R写R code

2. R in ipython/jupyter

# ipython magic method for writing R code in a cell# %%R must be put in the fist line in a cell%%Rcat('iris class: ', class(iris), '\n')print(head(iris))plot(iris)

3. Communacating between Python and R

3.1 Python to R

import pandas as pd
df = pd.DataFrame({'x':['a','b','c','a','b'],'y':[1,2,3,4,5]})print('df class: ', type(df))print(df)

# pushing pandas dataframe into R dataframe%Rpush df

/home/xiangjie/programs/anaconda3/envs/py36/lib/python3.6/site-packages/rpy2/robjects/pandas2ri.py:191: FutureWarning: from_items is deprecated. Please use DataFrame.from_dict(dict(items), ...) instead. DataFrame.from_dict(OrderedDict(items)) may be used to preserve the key order.
  res = PandasDataFrame.from_items(items)

%%Rlibrary('ggplot2')  # 用test环境（刚刚安装的这个conda环境）下的R安装一下就好（在teminal中conda activate test之后，再执行R就是这个环境下的R了）
cat('df class: ', class(df), '\n')


    
print(head(df))(ggplot(df, aes(x,y)) + geom_boxplot())# ggsave('./test.pdf')

/home/xiangjie/programs/anaconda3/envs/py36/lib/python3.6/site-packages/rpy2/rinterface/__init__.py:146: RRuntimeWarning: Want to understand how all the pieces fit together? See the R for Data
Science book: http://r4ds.had.co.nz/

  warnings.warn(x, RRuntimeWarning)

%%Rres summary(res)

3.2 R to Python

%%Rvec 1,2,3)cat('vec class:', class(vec), '\n')print(vec)

%Rpull vecvec_py = list(vec)print('vec class: ', type(vec_py))print(vec_py)

R to Python在数据形式复杂的时候比较麻烦，个人用的不多，具体可以查看rpy2的tutorial (见references)

In [11]:

## ipython还有很多魔法方法，其中比较常用的有!后直接跟shell命令来直接运行shell命令（比如!ls），%%time/%time来计算一段代码的运行时间等等
# %lsmagic  # 列出IPython的魔法方法

References:

rpy2: https://rpy2.readthedocs.io/en/version_2.8.x/

ipython rmagic extension: https://ipython.org/ipython-doc/2/config/extensions/rmagic.html

ipython rmagic extension: https://rpy2.github.io/doc/v2.9.x/html/interactive.html

ipykernel: https://ipython.readthedocs.io/en/stable/install/index.html