rpy2 : Korean characters are not working on rpy2

Question

python code:

import rpy2.robjects as robjects

rCommand='''gender <- c("남자", "남자", "남자", "여자", "여자", "여자", "여자", "여자")
  age    <- c(23, 25, 27, 29, 31, 33, 35, 37)
  outdf <- data.frame(gender, age)
'''
robjects.r(rCommand)
resultDf_r=robjects.globalenv["outdf"]

print type(resultDf_r)

Korean characters makes the python.exe killed.

In R command terminal, "rCommand" above works well.

I couldn't find any solution.

Any help would be appreciated.

my env: window 7 x64, python 2.7.8 x64, rpy2 2.5.4, R 3.1.2

Answer 1

At the time of writing, there is unfortunately no official support of rpy2 for windows. The code snippet you are providing is working fine on Linux.

Your options might be:

• run your code in a Linux VM (or container - MS has announced to be backing Docker)

• submit a patch for rpy2

• file a bug report on the bitbucket page for rpy2 hoping that this translates into a patch by someone

Edit: The comments suggest that helping out Python 2.7 with encoding might help (no promise - everything is working on Linux so this might be Windows-specific). A string can be explicitly specified to be in unicode with (note the prefix u before ''' ):

rCommand=u'''
    gender <- c("남자", "남자", "남자", "여자", "여자", "여자", "여자", "여자")
    age    <- c(23, 25, 27, 29, 31, 33, 35, 37)
    outdf <- data.frame(gender, age)
'''
robjects.r(rCommand)