CODEWITHSUNDEEP
pythonpandasdataframe
themli
themli

Reputation: 99

Pandas: Selecting Column based on row content

Suppose you have the following dataframe:

        Col1     Col2    Col3    Col4   Col5
Index
Row1    Code1   Code2   Code3  Code4  Code5
Row2    1       0       1       1       0
Row3    10      0       10      0       10
Row4    7070    56746   87647   969     687876
Row5    98798   79657   6876    977     6678

Is there a way to select columns based on criteria for the rows (index)?

So for instance,

for all columns with Row2==1:
    do something with all the rows in those columns

and similarly

for all columns with Row3==0:
    do something with all the rows in those columns

etc.

I was thinking something like 

for "Row3" in df.index:
    if "Row3" == 0:
        # Do certain calculation

Upvotes: 1

Views: 368

Answers (1)

ansev
ansev

Reputation: 30920

Use DataFrame.loc:

m2 = df.loc['Row2'].eq(1)
#m2 = df.loc['Row2'].eq('1') #if string type

cols_Row2 = df.columns[m2]
#Index(['Col1', 'Col3', 'Col4'], dtype='object')

#for col in cols_Rows2:
    #.....

Indexing the dataframe

df.loc[:,m2] 
#         Col1    Col3    Col4
#Index                        
#Row1    Code1   Code3   Code4 
#Row2        1       1       1
#Row3       10      10       0
#Row4     7070   87647     969
#Row5    98798    6876     977

Upvotes: 1

Related Questions