Can't set index of a pandas data frame - getting “KeyError”

Question

I generate a data frame that looks like this (summaryDF):

   accuracy        f1  precision    recall
0     0.494  0.722433   0.722433  0.722433
         


        

Answer 1

I guess you and @jezrael misunderstood an example from the pandas docs:

df.set_index(['A', 'B'])

A and B are column names / labels in this example:

In [55]: df = pd.DataFrame(np.random.randint(0, 10, (5,4)), columns=list('ABCD'))

In [56]: df
Out[56]:
   A  B  C  D
0  6  9  7  4
1  5  1  3  4
2  4  4  0  5
3  9  0  9  8
4  6  4  5  7

In [57]: df.set_index(['A','B'])
Out[57]:
     C  D
A B
6 9  7  4
5 1  3  4
4 4  0  5
9 0  9  8
6 4  5  7

The documentation says it should be list of column labels / arrays.

so you were looking for:

In [58]: df.set_index([['A','B','C','D','E']])
Out[58]:
   A  B  C  D
A  6  9  7  4
B  5  1  3  4
C  4  4  0  5
D  9  0  9  8
E  6  4  5  7

but as @jezrael has suggested df.index = ['A','B',...] is faster and more idiomatic method...