python - Nested dictionary from pandas data frame -

i have following data (see below) in pandas data frame.

i'd covert dict looks this:

my_dict = {  'ab': {          'city1': (0.000000, 0.000000),          'city2' : (0.100000, 0.200000),           'city3' : (0.200000, 0.400000)         } 'bc': {          'city4':  (0.300000, 0.600000),          'city5' : (0.400000, 0.800000),       }      } 

i aware of pandas' to_dict() method have been unable coerce perform this.

---

   iso                city        lat        lng 0   ab               city1  0.000000    0.000000 1   ab               city2  0.100000    0.200000 2   ab               city3  0.200000    0.400000 3   bc               city4  0.300000    0.600000 4   bc               city5  0.400000    0.800000 

you can first create column zipped zip lat , lng , groupby double to_dict:

#python 3 need convert list df['zipped'] = list(zip(df.lat, df.lng)) print (df)   iso   city  lat  lng      zipped 0  ab  city1  0.0  0.0  (0.0, 0.0) 1  ab  city2  0.1  0.2  (0.1, 0.2) 2  ab  city3  0.2  0.4  (0.2, 0.4) 3  bc  city4  0.3  0.6  (0.3, 0.6) 4  bc  city5  0.4  0.8  (0.4, 0.8)  d = df.groupby('iso').apply(lambda x: x.set_index('city')['zipped'].to_dict()).to_dict() print (d)  {'ab': {'city3': (0.20000000000000001, 0.40000000000000002),          'city1': (0.0, 0.0),          'city2': (0.10000000000000001, 0.20000000000000001)},  'bc': {'city4': (0.29999999999999999, 0.59999999999999998),         'city5': (0.40000000000000002, 0.80000000000000004)}}