Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Get Started

MLCast is split across several repositories under github.com/mlcast-community, built around three main pieces that follow the data from raw radar to trained model:

  1. Datasets (mlcast-datasets) — an Intake catalog of curated, mlcast-compliant source radar datasets, with per-provider converters and a sampler that turns source data into training-ready indices.

  2. Validator (mlcast-dataset-validator) — checks that a contributed dataset meets the MLCast Zarr specification and works with common geospatial tools before it joins the catalog.

  3. Machine learning (mlcast) — the main Python package: ConvGRU ensemble nowcasting models and a configuration-driven CLI / Python API for training.

The typical flow: a dataset is converted and validated, registered in the catalog, sampled into training indices, then fed to the mlcast package for training.

Quick start

  1. Get data — open the Intake catalog of curated radar datasets. Install from GitHub for the latest version, or pip install mlcast-datasets for a stable release. See Datasets.

  2. Validate (contributors only) — if you are adding a new dataset, check it against the MLCast Zarr spec first. See Validator.

  3. Train a model — fork and clone mlcast, install with uv sync, then run mlcast train. See Usage.