harrisonzhu508 / m4r_project

Source code for M4R Project

m4r_project: Source code for M4R Project

The minimal working example (MWE) is provided below.

Requirements

• Google Colab access or GPU-enabled machines
• Python 3.7 or above
• R 3.5.1 or above

Sweden Crop Yield Forecast

Simply run the following commands from the root m4r_project folder:

cd sweden_bart/R
./run_monthly.sh

This will produce plots and saved models in the sweden_bart/saved/ folder. Then open notebooks/sweden.ipynb and configure to run the XGB-GP models and full analysis.

USA Crop Yield Forecast

Simply run the following commands from the root m4r_project folder:

cd usa_bart/R
./run_monthly.sh

This will produce plots and saved models in the usa_bart/saved/ folder. Then open notebooks/usa.ipynb and configure to run the XGB-GP models and full analysis.

Other code

src/ contains code used for data engineering and collection.