Extracting first 3 elements from list of strings in pandas df

Question

I want to extract the first 3 elements from a list of strings from the '1/1' column. My df_unique looks like that:

                                                                                    1/1                                                             0/0  count
0              ['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-28', 'P1-6', 'P1-88', 'P1-93']                            ['P1-89', 'P1-90', 'P1-92', 'P1-95']      1
1              ['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-6', 'P1-89', 'P1-92', 'P1-95']                                     ['P1-28', 'P1-90', 'P1-93']      1
2    ['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-88', 'P1-89', 'P1-92', 'P1-93', 'P1-95']                                      ['P1-28', 'P1-6', 'P1-90']      1
3             ['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-88', 'P1-89', 'P1-92', 'P1-93']                                      ['P1-28', 'P1-6', 'P1-90']      1                                                                                                                                         

I've tried to use different solutions:

df_extract_3 = df_unique['1/1'].str.split().map(lambda lst: [string[0:3] for string in lst])

but the result looks like that:

0           [['P, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1]
1           [['P, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1]
2      [['P, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1]
3           [['P, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1, 'P1]

And the second solution:

df_extract_3 = df_unique['1/1'].str[0:3]

gives:

0      ['P
1      ['P
2      ['P
3      ['P

When I try to add split :

df_extract_3 = df_unique['1/1'].str.split().str[0:3]

the final result is:

0      [['P1-12',, 'P1-22',, 'P1-25',]
1      [['P1-12',, 'P1-22',, 'P1-25',]
2      [['P1-12',, 'P1-22',, 'P1-25',]
3      [['P1-12',, 'P1-22',, 'P1-25',]

What should I change to receive 'normal' output like:

0      ['P1-12', 'P1-22', 'P1-25']
1      ['P1-12', 'P1-22', 'P1-25']
2      ['P1-12', 'P1-22', 'P1-25']
3      ['P1-12', 'P1-22', 'P1-25']

? I know it can be easy modification but I've stuck and messed with that... Thanks a lot!

Answer 1

First convert your strings to read lists, then slice with str:

import ast
df_unique['1/1'] = df_unique['1/1'].apply(ast.literal_eval)
df_unique['0/0'] = df_unique['0/0'].apply(ast.literal_eval)

df_extract_3 = df_unique['1/1'].str[:3]
print(df_extract_3)

Or in one shot:

df_extract_3 = df_unique['1/1'].apply(lambda x: ast.literal_eval(x)[:3])

Output:

0    [P1-12, P1-22, P1-25]
1    [P1-12, P1-22, P1-25]
2    [P1-12, P1-22, P1-25]
3    [P1-12, P1-22, P1-25]
Name: 1/1, dtype: object

Answer 2

try this:

data={"1/1":[['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-28', 'P1-6', 'P1-88', 'P1-93'],
        ['P1-12', 'P1-22', 'P1-25', 'P1-26', 'P1-6', 'P1-89', 'P1-92', 'P1-95']],
"0/0":[1,2]}
df=pd.DataFrame(data)

df["1/1"]=df["1/1"].apply(lambda x:x[0:3])

print(df)
                     1/1    0/0
0   [P1-12, P1-22, P1-25]   1
1   [P1-12, P1-22, P1-25]   2