pump_null_zero_geographicals_1.py

# bound of min/max latitude/longitude/height for Tanzania
bound_df = dfX[(dfX['latitude']<-0.5)&(dfX['longitude']>25)&(dfX['gps_height']>0)]

# mean of geographical data in each bucket
mean_geo_df = bound_df.groupby(['basin',])['latitude','longitude','gps_height'].mean()

assert(mean_geo_df.shape[0] == len(dfX['basin'].unique()))

# Out[31]: mean_geo_df
#                           latitude  longitude gps_height
# basin
# Internal                  -4.1370   35.4715   1474.5504
# Lake Nyasa                -10.021   34.7726   1524.6542
# Lake Rukwa                -7.4896   31.5001   1415.1373
# Lake Tanganyika           -4.9386   30.3982   1270.0632
# Lake Victoria             -1.8684   34.0041   1329.2650
# Pangani                   -3.8493   37.5468   1102.2577
# Rufiji                    -8.4833   35.8520   1087.7806
# Ruvuma / Southern Coast   -10.547   38.1568   455.0583
# Wami / Ruvu               -6.7127   38.1703   308.9142