You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The corr method for DataFrames is exceedingly useful, especially since one can specify methods other than the Pearson correlation. I think it would be nice if this same argument were available when using corrwith. Perhaps there is another workaround, but right now if one wants to calculate all the Spearman (say) correlations with a specific variable, it seems necessary to calculate the full Spearman correlation matrix and subset to the column you care about.
Thanks.
The text was updated successfully, but these errors were encountered:
dsaxton
changed the title
Allow different methods of correlation when using corrwith
Allow different methods of correlation when using corrwith
Jul 15, 2018
Looks like the corr algorithms are written in Cython but corrwith uses a Python implementation of the Pearson calculation. Should be easy enough to add the other coefficient options to the corrwith method but they should probably use Cython implementations too since Spearman and Kendall are relatively complex calculations.
I can take a look at this and see if the algorithms that corr uses can also be reused by corrwith in some way, shouldn't be too tricky
Hi,
The
corr
method for DataFrames is exceedingly useful, especially since one can specify methods other than the Pearson correlation. I think it would be nice if this same argument were available when usingcorrwith
. Perhaps there is another workaround, but right now if one wants to calculate all the Spearman (say) correlations with a specific variable, it seems necessary to calculate the full Spearman correlation matrix and subset to the column you care about.Thanks.
The text was updated successfully, but these errors were encountered: