Solar Wind classification using Self-Organizing Maps.

This repository is under revision and does not contain the latest version of the code. The new version will be uploaded in the next days, please visit this page again soon, or contact the corresponding author for more details.

Please check the requirements.txt file for information about the packages and versions used to produce the publication results.

To install requirements use:

pip install -r requirements.txt

The main file controling the execution of the full architecturs is swinsom-ace.py. Please launch this file using the following arguments:

python swinsom-ace.py <case> <mode>

where <case> can be any of Amaya, Roberts, XuBorovsky or ZhaZuFi, and <mode> can be new or load.

The new option will launch the full data pre-processing, encoding, training and figure reproduction. The option load will load the data stored in the pickle files. Please check the optional flags inside the swinsom-ace.py file to change the run options. The ACE files are available via FTP in this link at the ACE Science Center. The data has been tranformed from .hdf (HDF4) format to .h5 (HDF5) format, using hdfview.

This code was used in the following publications:

• Special Issue of Frontiers in Astronomy and Space Sciences Space Physics.

The software in this repository has been written by Jorge Amaya