CODEWITHSUNDEEP
pythonpandasnumpy
Eran Moshe
Eran Moshe

Reputation: 3208

Pandas - revert many hot encoded (dummy variables)

Assuming I have the following pandas.DataFrame:

df = pd.DataFrame({'id': [1, 2, 3], 'val': [5, 5, 10],
                   'trig_aaa': [1, 0, 1], 'trig_bbb': [0, 1, 1], 'trig_ccc': [0, 0, 1]})
print(df)

   id  val  trig_aaa  trig_bbb  trig_ccc
0   1    5         1         0         0
1   2    5         0         1         0
2   3   10         1         1         1

I'd like to turn it to the following df:

   id  val             trig
0   1    5            [aaa]
1   2    5            [bbb]
2   3   10  [aaa, bbb, ccc]

is there an elegant (hopefully, functionality pre-built) in Pandas/Python/Numpy?

EDIT 1:

After looking at jpps' comment, a better processing to the DataFrame would looks like so:

   id  val trig
0   1    5  aaa
1   2    5  bbb
2   3   10  aaa
3   3   10  bbb
4   3   10  ccc

Upvotes: 1

Views: 260

Answers (1)

jpp
jpp

Reputation: 164843

You can use pd.melt:

# rename columns and melt dataframe
df.columns = [i if '_' not in i else i.split('_')[1] for i in df]
res = pd.melt(df, id_vars=['id', 'val'], var_name='trig')

# filter for 1 values and sort
res = res[res['value'].eq(1)].sort_values('id').iloc[:, :-1].reset_index(drop=True)

print(res)

   id  val trig
0   1    5  aaa
1   2    5  bbb
2   3   10  aaa
3   3   10  bbb
4   3   10  ccc

Upvotes: 2

Related Questions