Spy4Cast is an open-source python implementation of a maximum convariance analysis (MCA) statistical model.
In this manual we explain the steps to take to install and run for the first time this model.
Full documentation is available at https://spy4cast-docs.netlify.app
Link to GitHub repository with the latest code: https://github.com/pabloduran016/Spy4CastManual
You have to have installed git (Install Git https://git-scm.com/book/en/v2/Getting-Started-Installing-Git)
You have to have installed anaconda (Install Anaconda https://docs.anaconda.com/anaconda/install/index.html)
$ conda create -n <your-env-name> python=3.9
$ conda activate <your-env-name>
(<your-env-name>) $ conda install pip
(<your-env-name>) $ conda install cartopy
(<your-env-name>) $ pip install git+https://github.com/pabloduran016/Spy4Cast
Cartopy has to be installed with conda because pip version does not work
For usage example look at file EquatorialAtlantic_Impact_Nino.py in this repository.
DOI: https://10.5281/zenodo.8162066
BibTex:
@software{duran_fonseca_pablo_2023_8219489,
author = {Duran-Fonseca, Pablo and
Rodriguez-Fonseca, Belen},
title = {Spy4CastManual},
month = aug,
year = 2023,
publisher = {Zenodo},
version = {1.0.0},
doi = {10.5281/zenodo.8219489},
url = {https://doi.org/10.5281/zenodo.8219489}
}