/Hybrid-Forecasting-Competition

Notebooks Created as part of the Hybrid Forecasting Competition

Primary LanguageJupyter Notebook

README
Here are some of the notebooks in the Hybrid Forecasting Competition.

0. Create a dir data0928 under the current dir, and move forecasts.json and ifps.json here, renaming as forecasts0928.json and ifps0928.json.

1. First run translate_ifp_fcast_jsons_into_csvs(Sep2018).ipynb to generate several preprocessed csv from forecasts.json and ifps.json
2. Then run preprocessing_including_unresolved_IFPs(Sep2018).ipynb to generate data matrices as .npy file.
3. Then run SBS_BS_multiplicative_weights_aggregation(Sep2018).ipynb to run aggregators with input data stored in the .npy just generated.

Let me know if there are any questions.