Cross-country final model training

[tm-jace-peralta]

Adds the final cross-country model training notebook.

• The notebooks and output models are patterned after the single country training notebooks from https://github.com/thinkingmachines/unicef-ai4d-poverty-mapping/pull/150
• Adds the final model file which will be used in rollout (model_cross_country_rollout.pkl)
• Modified feature_engineering.py to add a function scale_dataframe_column that enables sklearn scalers to be used in pandas groupby functions

Check: 2023_02-21-crosscountry_final_model_training.ipynb

• Cell [5] with the loop to load, scale and concatenate the cross-country dataframe
• Cell [9] on implementation of scaling within groupby

scale_dataframe_column

• Lines [126-139] for implementation of scale_dataframe_column

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[alronlam]

Did some fixes and clean-ups. Specifically:

• Fixed non-removal of unscaled feature columns
• Removed unnecessary sections (evaluation; do not want to confuse people with two sets of evaluation metrics. our original leave one country out cross val metrics should suffice already)
• Updated wrongly copy-pasted notes (some of the notes still indicated PH)
• Re-organized sections a bit (made the notebook less verbose so it's easier to follow the actual code to generate the data and produce a model)

[tm-jc-nacpil]

@alronlam @tm-jace-peralta LGTM! On a side note, should we also make the single-country training notebooks less verbose too?