Tutorials

This repository includes my tutorials for data-driven methods for climate science applications. You can find the following tutorials:

1. AE_VAE/: Tutorial on autoencoders and Variational Autoencoders for SSTA in the tropical Pacific and SLP over the Atlantic

   

2. MLP_timeseries/: Introduction to multi-layer perceprons for time-series forecasting.

   

3. EOF/: Empirical orthogonal function analysis (PCA for spatio-temporal data).

   

4. LIM/: Linear inverse model (Dynamical mode decomposition) for SSTA field prediction.

   

Installation

Due to dependencies I recommend using conda. A list of packages is provided in the 'condaEnv.yml' file. The following steps set up a new environment with all required packages:

1. Install packages:

conda env create -f condaEnv.yml

2. Activate environment:

conda activate tutorialEnv

Note: If you don't have a Nvidia grafic card you need to comment the line including cudatoolkit in the condaenv.yml file