This project uses conda to manage dependencies. You can use the following command to install the required dependencies:
conda create -n tc -f environment.yml
After the installation finishes,
you can activate the environment tc to run the experiments in this project.
This project utilizes Jupytext to save jupyter notebooks as normal python scripts.
Therefore, in order to run the experiments as jupyter notebooks, you have to open these python scripts as jupyter notebooks. You can follow the instruction here to do that.
The project is organized as followed:
ncep/folder contains python scripts and bash scripts to process NCEP/FNL data into binary datasets that can be consumed by tensorflow.