Due to work needs, in doing the optimal split box, always write out the efficient code, so I found the R language in the optimal split box package, this time taking into account the call in python R language, the perfect combination. In the domestic Chinese site searched half a day, build the environment has been the emergence of " R_user not defined "this error, simply annoying, and finally on the site to find a solution.
My environment is: win7 ;
Use anaconda3 ;
(i) First you need an anaconda package;
Not much more to say on this one.
(ii) You need to have an R program.
I use the Microsoft development of this. Of course you use the R language software from the R official website, or you use RStudio is not a big problem.
/download
Setting environment variables to complete R with
1. Add the R directory to the path environment variable later.
path：D:\R\microsoft-r-open-3.4.2\bin\x64;
2. Add to the path environment variable
path：D:\R\microsoft-r-open-3.4.2\bin\x64\;
(c) Download rpy2 based on the experience of a small group of friends
Going to go unofficial and download a whl file.
/~gohlke/pythonlibs/#rpy2
Please recognize the python version and the 64-bit download, after downloading, install , put it under D:\Anaconda3\Anaconda3\Lib\site-packages, anyway, under your anaconda's site-packages.
command line in cmd, and then go to the path you just took.
C:\Users\Administrator>D:
D:\>cd D:\Anaconda3\Anaconda3\Lib\site-packa
D:\Anaconda3\Anaconda3\Lib\site-packages>pip install rpy2-2.8.6-cp36-cp36m-win_amd64.whl
At the end it will say successful and you are successful;
(iv) Configure the environment to test whether the code can run with R
Open up your python and bang on this code as you see fit and try it out.
In [1]: import os
In [2]: ['R_HOME'] = 'D:\R\microsoft-r-open-3.4.2'
In [3]: ['R_USER'] = 'D:\Anaconda3\Anaconda3\Lib\site-packages\rpy2'
In [4]: import as robjects
In [5]: ('''
...: # create a function `f`
...: f <- function(r, verbose=FALSE) {
...: if (verbose) {
...: cat("I am calling f().\n")
...: }
...: 2 * pi * r
...: }
...: # call the function `f` with argument value 3
...: f(3)
...: ''')
Out[5]:
R object with classes: ('numeric',) mapped to:
<FloatVector - Python:0x000000000B06FD88 / R:0x000000000EFE65B8>
[18.849556]
(v) Calling R's package
We usually use python to call R, and we usually have to use a special package that's not in python or something like that. So now I'm going to call the smbinning package, which is a package that enables optimal segmentation of data when it's split into boxes.
In [6]: from import importr
In [7]: smbinning = importr('smbinning')
This is the entire content of this article.