Wsrtjtyk

I have a problem with a data type.

Please assumed that here is my sample data frame.

class1 class2 docid

 A123   08/9   X123

 A123   08/1   X123

 A124   08/1   X124

 A124   08/2   X124

 A125   08/3   X125

I have merged class1 and class2 then named as class3

  class3     docid

 A123,08/9   X123

 A123,08/1   X123

 A124,08/1   X124

 A124,08/2   X124

 A125,08/3   X125

and then make a matrix by get_dummies

df1 = pd.get_dummies(df.docid).sort_index(level=0).max(level=[0,1])

df1

and get the results like this

class3     X123 X124 X125

A123,08/9   1    0    0

A123,08/1   1    0    0

A124,08/1   0    1    0

A124,08/2   0    1    0

A125,08/3.  0    0    1

and then I have dropped the class3 then transposed this matrix to calculate the Jaccard similarity by docid

df1_new = df1.drop(['class3'], axis=1)

df1_new_1 = df1_new.transpose()

df1_new_1

and the results are being like this

     0 1 2 3 4 

X123 1 1 0 0 0

X124 0 0 1 1 0

X125 0 0 0 0 1

from this result, the column has no name, then I would like to ask how can I change the X123 X124 X125 into 0 1 2 or only change the datatype from string to int? because, when I use this result to calculate the Jaccard similarity, it's appeared

ValueError: invalid literal for int() with base 10: 'X123'

thank you in advance

if your problem is just to change the str index to int , you can use

df1_new_1.reset_index(drop=True)

if you want to keep the values from string, you can use

df1_new_1.index.str.extract(r'd+')

Out:

Int64Index([123, 124, 125], dtype='int64')