Complex map function from a dictionary to generate two pandas columns

user_12751453 0
Viewed 36

I have the following dataframe:

{'POINT_ID': {0: 'ABC B4-14 c1-1', 1: 'ABC B4-14 c1-2', 2: 'ABC 14-10 c6-2', 3: 'ABC 14-10 c6-3', 4: 'ABC 14-03 c12-1', 5: 'ABC 14-03 c12-2', 6: 'ABC 14-01A c9-2', 7: 'ABC 14-01A c9-3', 8: 'DE 368 c1-1', 9: 'DE 368 c1-2', 10: 'DE 483 c3-2', 11: 'DE 483 c3-3', 12: 'FG 1 c4-1', 13: 'HI 1A c4-3'}, 'Count': {0: 1.25, 1: 2.66, 2: 3.5, 3: 6.6, 4: 9.54, 5: 10.57, 6: 11.8, 7: 2.5, 8: 1.1, 9: 1.04, 10: 2.22, 11: 3.22, 12: 1.5, 13: 1.2}}

I want to map the values in the column POINT_ID, using this dictionary:

labels = {'ABC B4-14': ['Local 1', 'Context 1'],
          'ABC 14-10': ['Local 2', 'Context 2'],
          'ABC 14-03': ['Local 2', 'Context 2'],
          'ABC 14-01A': ['Local 1', 'Context 1'],
          'DE 368': ['Local 3', 'Context 3'],
          'DE 483': ['Local 3', 'Context 4'],
          'FG 1': ['Local 4', 'Context 5'],
          'HI 1A': ['Local 5', 'Context 6']}

So I can generate the following dataframe:

{'POINT_ID': {0: 'ABC B4-14 c1-1', 1: 'ABC B4-14 c1-2', 2: 'ABC 14-10 c6-2', 3: 'ABC 14-10 c6-3', 4: 'ABC 14-03 c12-1', 5: 'ABC 14-03 c12-2', 6: 'ABC 14-01A c9-2', 7: 'ABC 14-01A c9-3', 8: 'DE 368 c1-1', 9: 'DE 368 c1-2', 10: 'DE 483 c3-2', 11: 'DE 483 c3-3', 12: 'FG 1 c4-1', 13: 'HI 1A c4-3'}, 'Count': {0: 1.25, 1: 2.66, 2: 3.5, 3: 6.6, 4: 9.54, 5: 10.57, 6: 11.8, 7: 2.5, 8: 1.1, 9: 1.04, 10: 2.22, 11: 3.22, 12: 1.5, 13: 1.2}, 'Local': {0: 'Local 1', 1: 'Local 1', 2: 'Local 2', 3: 'Local 2', 4: 'Local 2', 5: 'Local 2', 6: 'Local 1', 7: 'Local 1', 8: 'Local 3', 9: 'Local 3', 10: 'Local 3', 11: 'Local 3', 12: 'Local 4', 13: 'Local 5'}, 'Context': {0: 'Context 1', 1: 'Context 1', 2: 'Context 2', 3: 'Context 2', 4: 'Context 2', 5: 'Context 2', 6: 'Context 1', 7: 'Context 1', 8: 'Context 3', 9: 'Context 3', 10: 'Context 4', 11: 'Context 4', 12: 'Context 5', 13: 'Context 6'}}

The dictionary map keys are contained within the POINT_ID values, but they don't match exactly. Also, from the dictionary map values, there is a list and each member of the list needs to generate a different column.

pandas dictionary python
2 Answers

You could craft a DataFrame from the dictionary and merge on the extracted first part of POINT_ID:

df2 = pd.DataFrame(labels).set_axis(['Local', 'Context']).T
ID = df['POINT_ID'].str.extract('^(.*?) [\S]+$', expand=False)
out = df.merge(df2, left_on=ID, right_index=True).drop(columns='key_0')

Other option, map and join:

ID = df['POINT_ID'].str.extract('^(.*?) [\S]+$', expand=False)
out = df.join(pd.DataFrame(ID.map(labels).to_list(), columns=['Local', 'Context']))

output:

           POINT_ID  Count    Local    Context
0    ABC B4-14 c1-1   1.25  Local 1  Context 1
1    ABC B4-14 c1-2   2.66  Local 1  Context 1
2    ABC 14-10 c6-2   3.50  Local 2  Context 2
3    ABC 14-10 c6-3   6.60  Local 2  Context 2
4   ABC 14-03 c12-1   9.54  Local 2  Context 2
5   ABC 14-03 c12-2  10.57  Local 2  Context 2
6   ABC 14-01A c9-2  11.80  Local 1  Context 1
7   ABC 14-01A c9-3   2.50  Local 1  Context 1
8       DE 368 c1-1   1.10  Local 3  Context 3
9       DE 368 c1-2   1.04  Local 3  Context 3
10      DE 483 c3-2   2.22  Local 3  Context 4
11      DE 483 c3-3   3.22  Local 3  Context 4
12        FG 1 c4-1   1.50  Local 4  Context 5
13       HI 1A c4-3   1.20  Local 5  Context 6
user_163434640

I'm assuming I'm taking everything up to the last space.

df2 = pd.DataFrame(
    df.POINT_ID.str.rsplit(n=1).str[0].map(labels).tolist(),  # Where magic happens
    index=df.index, columns=['Local', 'Context']
)
df.join(df2)

           POINT_ID  Count    Local    Context
0    ABC B4-14 c1-1   1.25  Local 1  Context 1
1    ABC B4-14 c1-2   2.66  Local 1  Context 1
2    ABC 14-10 c6-2   3.50  Local 2  Context 2
3    ABC 14-10 c6-3   6.60  Local 2  Context 2
4   ABC 14-03 c12-1   9.54  Local 2  Context 2
5   ABC 14-03 c12-2  10.57  Local 2  Context 2
6   ABC 14-01A c9-2  11.80  Local 1  Context 1
7   ABC 14-01A c9-3   2.50  Local 1  Context 1
8       DE 368 c1-1   1.10  Local 3  Context 3
9       DE 368 c1-2   1.04  Local 3  Context 3
10      DE 483 c3-2   2.22  Local 3  Context 4
11      DE 483 c3-3   3.22  Local 3  Context 4
12        FG 1 c4-1   1.50  Local 4  Context 5
13       HI 1A c4-3   1.20  Local 5  Context 6
user_23366540
Related