Metadata-Version: 2.3
Name: flarenet20
Version: 0.1.1
Summary: a convolutional neural network to predict flares in 20-second TESS data
License: MIT
Author: Vera Berger
Author-email: vlberger@mit.edu>, Nicole Schanche <nicole.e.schanche@nasa.gov
Requires-Python: >=3.9,<=3.12
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Dist: astropy (>=5.3)
Requires-Dist: batman-package (>=2.4.8,<3.0.0)
Requires-Dist: lightkurve (>=2.5.0)
Requires-Dist: lksearch (>=1.2.0)
Requires-Dist: numpy (>=1.26.0)
Requires-Dist: pandas (>=2.2,<3.0)
Requires-Dist: scikit-learn (>=1.3.0)
Requires-Dist: tensorflow (>=2.17.0,<2.18.0)
Description-Content-Type: text/markdown

**flarenet** is a **convolutional neural network** used to **predict flares** in 20-second cadence data from NASA's Transiting Exoplanet Survey Satellite (**TESS**).

With a single line of code, you can generate flare predictions for any TESS 20-s target.

![Output predictions](https://github.com/veraberger/flarenet/blob/5f86206003fa66fc4a9390170b6ac45fdc9dfa39/figures/flare_predictions.png)

### Installation
With pip:
```bash
pip install flarenet20
```

For developers:
```bash
git clone https://github.com/veraberger/flarenet.git
cd flarenet
pip install .
```



### Usage
Tutorial notebook: [docs/example.ipynb](docs/example.ipynb)

Use our model trained on real TESS light curves with injected false-positives, or train your own with the ```Flarenet``` class.

The model outputs likelihoods only - choose your own confidence threshold for flares.


If you use this code in publications, please cite Berger, Schanche, et al. (2026).

MIT License

