Unable to rename column in dataframe

Question

I am trying to rename the '0' to 'predicted' as per below I receive no error but when I print it, it still shows as '0'?

df2 = pd.DataFrame(data)
df2.rename(columns = {'0':'predicted'}, inplace=True)
print (df2.tail())

            0
248  0.335400
249  0.334992
250  0.334955
251  0.335716
252  0.335723

Answer 1

I believe 0 is integer, so change '0' (string) to 0 (number):

df2.rename(columns = {0:'predicted'}, inplace=True)

But better is use parameter columns in DataFrame constructor:

df2 = pd.DataFrame(data, columns=['predicted'])

Answer 2

Here is another generic solution which should work for both - columns as numbers and columns as string representation of numbers:

Sample DF:

In [32]: df = pd.DataFrame(np.arange(15).reshape(-1,5), columns=[0,'1',2,'colX','colZ'])

In [33]: df
Out[33]:
    0   1   2  colX  colZ
0   0   1   2     3     4
1   5   6   7     8     9
2  10  11  12    13    14

In [34]: df.columns.tolist()
Out[34]: [0, '1', 2, 'colX', 'colZ']

Mapping:

In [29]: d = {'0':'col0','1':'col1','2':'col2'}

Renaming:

In [35]: df = df.rename(columns=lambda col: d.get(str(col)) if str(col) in d else col)

In [36]: df
Out[36]:
   col0  col1  col2  colX  colZ
0     0     1     2     3     4
1     5     6     7     8     9
2    10    11    12    13    14