User guide
Contents:
Installation
To use cloudcasting, first install it using pip:
git clone https://github.com/alan-turing-institute/cloudcasting
cd cloudcasting
python -m pip install .
Optional dependencies
cloudcasting supports optional dependencies, which are not installed by default. These dependencies are required for certain functionality.
To run the metrics on GPU:
python -m pip install --upgrade "jax[cuda12]"
To make changes to the library, it is necessary to install the extra dev dependencies, and install pre-commit:
python -m pip install ".[dev]"
pre-commit install
To create the documentation, it is necessary to install the extra doc dependencies:
python -m pip install ".[doc]"
Getting started
Use the command line interface to download data:
cloudcasting download "2020-06-01 00:00" "2020-06-30 23:55" "path/to/data/save/dir"
Once you have developed a model, you can also validate the model, calculating a set of metrics with a standard dataset. To make use of the cli tool, use the model github repo template to structure it correctly for validation.
cloudcasting validate "path/to/config/file.yml" "path/to/model/file.py"