1

If I make a multi-indexed column dataframe like this:

iterables = [['bar', 'baz', 'foo', 'qux'], ['one', 'two']]
index = pd.MultiIndex.from_product(iterables, names=['first', 'second'])
df = pd.DataFrame(np.random.randn(3, 8), index=['A', 'B', 'C'], columns=index)


first        bar                 baz                 foo                 qux  \
second       one       two       one       two       one       two       one   
A      -0.119687 -0.518318  0.113920 -1.028505  1.106375 -1.020139 -0.039300   
B       0.123480 -2.091120  0.464597 -0.147211 -0.489895 -1.090659 -0.592679   
C      -1.174376  0.282011 -0.197658 -0.030751  0.117374  1.591109  0.796908   

first             
second       two  
A      -0.938209  
B      -0.851483  
C       0.442621

and I want to select columns from only the first set of columns using a list,

select_cols=['bar', 'qux']

such that the result would be:

first        bar                  qux  
second       one       two        one        two
A      -0.119687 -0.518318  -0.039300  -0.938209    
B       0.123480 -2.091120  -0.592679  -0.851483    
C      -1.174376  0.282011   0.796908   0.442621

How would I go about doing that? (Thanks ahead of time)

Improve this question

3 Answers 3

Reset to default
5

Simple column selection works as well:

df[['bar', 'qux']]

# first        bar                 qux          
# second       one       two       one       two
# A       0.651522  0.480115 -2.924574  0.616674
# B      -0.395988  0.001643  0.358048  0.022727
# C      -0.317829  1.400970 -0.773148  1.549135
Improve this answer
Sign up to request clarification or add additional context in comments.

Comments

4

You can use loc to select columns:

df.loc[:, ["bar", "qux"]]

#  first       bar                    qux
# second       one        two         one         two
#      A  1.245525  -1.469999   -0.399174    0.017094
#      B -0.242284   0.835131   -0.400847   -0.344612
#      C -1.067006  -1.880113   -0.516234   -0.410847
Improve this answer

4 Comments

@iparjono More reasons or demonstrations about why it's not working?
I got the error KeyError: "[['bar', 'qux']] are not in ALL in the [columns]
Works for me on v0.18.1.
@iparjono It might be a version problem. I am on 0.18.1 too and it works.
2

When I found this Q/A I thought I might see a solution that prints the column names. Having figured it out, I thought I might add to the answer. The following prints out the values of the column name for a given level.

df.columns.get_level_values(0)

=> ['bar', 'qux']

- E

Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.