Calculate the number of occurrences of a column value value *per* unique id

Question

I want to calculate the number of occurrences of a columnvalue per unique id and store them in a new panda Dataframe:

Dataframe (simplified):

Desired output:

Keywords are +1.000s and not known. Have tried by creating lists and zipping them, but not really working.

Answer 1

尝试使用crosstab

pd.crosstab(df['unique id'],df['keyword'])

Answer 2

Use groupby apply with Counter , untack and fillna :

from collections import Counter
df.groupby('unique id')['keyword'].apply(Counter).to_frame().unstack(1).fillna(0)

          keyword            
             auto fish mobile
unique id                    
1a            2.0  1.0    1.0
2a            1.0  2.0    1.0
3a            0.0  1.0    0.0