在python中使用R脚本，这是为R编写的

Question

A friend has some R-scripts that I might find useful. But I use Python, and when he upgrades his scripts I want to be able to use his updates.

Is it possible to embed R-scripts as-is in Python?

An typical R-script he might write is named eg quadro.R and has the form:

quadro <- function(x) {
  return(x*x)}

Can I somehow call quadro.R from python with the argument "3" and get the result "9" back in Python? I do have R installed on my Linux system.

As I understand rpy/rpy2 , I can use R-commands in python but not use an R-script, or did I misunderstood something? Is there some other way to use an R-script from within Python?

Answer 1

First load the whole R script in python, then get any of its R object (function, variable, etc.) assigned and called in python.

An example python script,

from rpy2 import robjects

robjects.r('''                         
source('quadro.R')
''')                                   #load the R script

quadro = robjects.globalenv['quadro']  #assign an R function in python
quadro(3)                              #to call in python, which returns a list
quadro(3)[0]                           #to get the first element: 9

Answer 2

Rpy2 covers this kind of use rather well, I think. You can encapsulate loose R scripts into packages (and avoid storing objects into R's global environment - something that should be as carefully considered as the use of global variables in Python).

import rpy2.robjects.packages.SignatureTranslatedAnonymousPackage as STAP

with open('quadro.R') as fh:
    rcode = fh.read()
quadro = STAP(rcode, 'quadro')

# The function is now at
quadro.quadro()
# other functions or objects defined in that R script will also be there, for example
# as quadro.foo()

This is in the rpy2 documentation .