PyEarthTools is a Python framework, containing modules for loading data; pre-processing, normalising and standardising data; defining machine learning (ML) models; training ML models; performing inference with ML models; and evaluating ML models. It contains specialised support for weather and climate data sources and models. It has an emphasis on reproducibility, shareable pipelines, and human-readable low-code pipeline definition.
Note
THIS REPOSITORY IS UNDER CONSTRUCTION
This repository contains code which is under construction, and should not yet be used outside of a research setting. The development team are working busily to bring everything up to spec. As such, things are likely to change pretty often. Please take a look around!
We have information for:
- New user guides and introduction to the concepts in PyEarthTools
- Detailed Installation instructions (it's not hard, there are just lots of user environments to cater for)
- Data catalogue setup for facility managers or individuals to establish their research data catalogue
- A shiny tutorial gallery full of neat examples
- Much more, including how-to guides, project setup guide, information on accessing data, guides to evaluation, orientiation for physical scientists and data scientists at our documentation homepage (you may be reading this now or you may be visiting the README from elsewhere)