https://github.com/liamtoney/yasur_ml
Support vector classification of infrasound signals at Yasur Volcano
https://github.com/liamtoney/yasur_ml
infrasound machine-learning svm volcano
Last synced: 4 days ago
JSON representation
Support vector classification of infrasound signals at Yasur Volcano
- Host: GitHub
- URL: https://github.com/liamtoney/yasur_ml
- Owner: liamtoney
- License: mit
- Created: 2020-10-06T17:18:48.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2025-06-19T21:01:48.000Z (4 months ago)
- Last Synced: 2025-06-19T22:18:47.234Z (4 months ago)
- Topics: infrasound, machine-learning, svm, volcano
- Language: Python
- Homepage: https://doi.org/10.1785/0320220019
- Size: 266 KB
- Stars: 4
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
# yasur_ml
This repository contains the code accompanying the paper "Waveform features
strongly control subcrater classification performance for a large, labeled volcano infrasound
dataset" by [Liam Toney](mailto:ldtoney@alaska.edu), David Fee, Alex Witsil, and Robin S. Matoza.
## Installing
A conda environment specification file, [`environment.yml`](environment.yml), is
provided. You can create a conda environment from this file by executing
```shell
conda env create
```
from the repository root.
This code requires the [UAF Geophysics Tools](https://github.com/uafgeotools) package
[*rtm*](https://github.com/uafgeotools/rtm) for generation of the labeled
catalog. This package and its dependencies are installed when the above command
is executed.
You must define two environment variables to use the code:
- `YASUR_WORKING_DIR` — the path to this repository
- `YASUR_FIGURE_DIR` — the directory where figure files should be saved
## Workflow overview
1. [`download_3E.py`](data/download_3E.py) — download the data
2. [`build_catalog.py`](label/build_catalog.py) — run *rtm* to create a CSV catalog
3. [`label_catalog.py`](label/label_catalog.py) — associate entries in catalog to a subcrater
4. [`extract_features.py`](features/extract_features.py) — extract features from waveforms and store in Feather file
5. Apply tools in [`svm/`](svm/)
## Citation
If you use the tools contained in this repository, please cite our paper:
> Toney, L., Fee, D., Witsil, A., & Matoza, R. S. (2022). Waveform features
> strongly control subcrater classification performance for a large, labeled
> volcano infrasound dataset. *The Seismic Record*, *2*(3), 167–175.
> https://doi.org/10.1785/0320220019
## Acknowledgements
This work was supported by the Nuclear Arms Control Technology (NACT) program at the
Defense Threat Reduction Agency (DTRA). Cleared for release.