Filter pandas dataframe columns based on other dataframe

Question

I have two dataframes df1 and df2. df1 gives some numerical data on some elements (A,B,C ...) while df2 is a dataframe acting like a classification table with its index being the column names of df1. I would like to filter df1 by only keeping columns that are matching a certain classification in df2.

For instance, let's assume the following two dataframes and that I only want to keep elements (ie columns of df1) that belong to class 'C1':

df1 = pd.DataFrame({'A': [1,2],'B': [3,4],'C': [5,6]},index=[0, 1])

df2 = pd.DataFrame({'Name': ['A','B','C'],'Class': ['C1','C1','C2'],'Subclass': [C11,C12,C21]},index=[0, 1, 2])

df2 = df2.set_index('Name')

The expected result should be the dataframe df1 with only columns A and B because in df2, we can see that A and B are in class C1. Not sure how to do that. I was thinking about first filtering df2 by 'C1' values in its 'Class' column and then check if df1.columns are in df2.index but I suppose there is a much efficient way to do that. Thanks for your help

Answer 1

Here is one way using index slice

df1.loc[:,df2.index[df2.Class=='C1']]
Out[578]: 
Name  A  B
0     1  3
1     2  4