julianconstantine
commented
7 years ago Right now the list is the following:
- big_market_raion
- big_road1_1line
- culture_objects_top_25
- detention_facility_raion
- ecology
- incineration_raion
- nuclear_reactor_raion
- oil_chemistry_raion
- product_type
- radiation_raion
- railroad_1line
- railroad_terminal_raion
- sub_area
- thermal_power_plant_raion
- water_1line
- child_on_acc_pre_school
- modern_education_share
- old_education_build_share
When you run pandas's get_dummies() function (e.g. in my simple_xgb.py script), you need to specify which variables get binarized/turned into dummies. Otherwise, it will try to do this automatically, which probably isn't the best idea with ~300 features.
Also make sure to set drop_first to True next time you edit this.