xCDAT Validation

This repository is dedicated to xCDAT feature prototyping and validation. It is utilized in conjunction with the pull requests in the main repository here. Validation consists of Python scripts and Jupyter Notebooks, including comparisons of xarray/xcdat against a subset of CDAT functionality via cdms2, cdutil, and genutil.

Getting Started

Directory Layout

• scripts/ stores utility scripts
• validation/ stores validation scripts and notebooks

Setup

1. Clone this repo

   git clone https://github.com/xCDAT/xcdat_test

2. Create and activate a conda environment that includes xcdat, cdms2, cdutil, and genutil.

   • Work with the latest stable version of xcdat from conda-forge

     cd xcdat_test
     conda env create -f conda-env/test_stable.yml
     conda activate xcdat_test_stable

   • Work with the latest version of xcdat on the main branch

     cd xcdat_test
     conda env create -f conda-env/test_dev.yml
     conda activate xcdat_test_dev
     
     # Clone the latest version of xcdat and install into the `xcdat_test_dev` env
     git clone https://github.com/xCDAT/xcdat
     cd xcdat
     python -m pip install .

3. Start working with the Jupyter Notebooks!

Demo input preparation

Sample PMP download data for demos

wget https://raw.githubusercontent.com/PCMDI/pcmdi_metrics/master/doc/jupyter/Demo/Demo_0_download_data.ipynb

FYI: Useful External Resources

• xCDAT Docs

• Xarray Docs

• Xarray usage examples

  • A Quick Introduction to CMIP6
  • Creating a subset of dataset using Xarray
  • Xarray Access CMIP5 Data (NCI data training)
  • Visualize Climate data with Python

• Visualization examples

  • Hawkins Warming Stripes
  • Map using proplot
  • Map using cartopy