BasisResearch / cities

Home of Basis development for the 2023 TOP Sprint
MIT License
6 stars 0 forks source link

prep, model and train the interactions model #78

Closed rfl-urbaniak closed 10 months ago

rfl-urbaniak commented 10 months ago

Add the first version of prep_data_for_interaction_inference() (will be revised once we have invertible transformation in place).

Add train_interactions_model().

Add cities_model_interactions().

Why? First step towards building the guide-generating model training pipeline.

A brief test is in the experimental notebook named interaction_cities_prototype.ipynb. Ignore all other new stuff in the experimental notebooks folder.