Search code examples
pythonrrpy2

How to return multiple objects from R script executed in Python

So I am struggling to use the rpy2 package to integrate some workflow between R and Python.

For example, imagine I'm trying to run a Python script that does a linear regression in R and I want to return all of the elements of that (in reality I'm trying to do something much more complicated than this).

I execute the following in Python (calling R; This assumes you have rpy2 installed):

import rpy2.robjects as ro
test = ro.r('''
            # Load in data
            df <- mtcars
            # Run regression
            out = lm(formula='mpg ~ cyl + hp + wt',data=df)
            ''')

And now what? I have a questions:

  1. How do I pull the various elements from the result? In R they would be out$coefficients and out$residuals, etc. I know there is documentation for this, but I'm a bit lost. Ideally, I would want the elements in useful formats, so pandas dataframes or indexed lists, etc.

  2. What happens to df? robjects.r() seems to just save whatever the last thing you gave and throw away everything else. I suppose I can work with this, but it's not ideal.

  3. Related to 2: Is there a much much better way to do this? In general if someone could put forward a "best practice" for this sort of thing, that would be helpful, since I'm sure that there are many people interested in using Python, but occasionally have a very custom function they need to call using R but they don't want to get to fancy with the integration. Perhaps a way to call an R function using Pythonic input arguments would be great.

Solution

  • Q.1: How do I pull the various elements from the result?

    Ans.1: After you run your R script:

    test = ro.r(your_R_script)

    You can use this code to print out all the names and values in the test object.

    # iterate on names and values
# be careful output is v long
for n,v in test.items():
    print(n)
    print(v)

    To list all available names, run this code:

    test.names

    The output:

    StrVector with 12 elements.
'coeffici... 'residuals' 'effects' 'rank' ... 'xlevels' 'call'  'terms' 'model'

    To print values of the 'residuals', run this:

    test[test.names.index('residuals')]

    Q.2: What happens to df?

    Ans.2: It is still available in R environment until you delete it. You can run simple R code to check:

    ro.r('''
        # View dataframe
        df
        ''')

    Q.3: Is there a much much better way to do this?

    Ans.3: (No answer.)