The computing codes for the manuscript entitled "FUNCTIONAL NONLINEAR LEARNING"

The folder contains files to run the computing codes on the starlight data.

To run the python script: python starlight_compile.py StarLightCurves_TRAIN.tsv StarLightCurves_TEST.txv

Required libraries:

numpy, sklearn, tensorflow

Note that we need to use the version python 3.10.2 and tensorflow 2.8.0

Below are the discription of each file:

• starlight_compile.py: The main python script for running FunNoL. It produces a prediction accuracy on curve labels in test data.
• Starlight_model_compile.h5: a pre-trained FunNoL model.
• StarlightCurves_TRAIN.tsv: Training data
• StarlightCurves_TEST.tsv: Test data