Welcome to Deep-LC’s documentation!

Deep-LC is open-source and intended for the classification of light curves (LCs) in a gernaral purpose. It utilizes a weakly supervised object detection algorithm to automatically zoom in on the LC and power spectrum (PS) to extract local features. This eliminates the need for manual feature extraction and allows it to be applied to both space- and ground-based observations, as well as multiband LCs with large gaps and nonuniform sampling.

It can zooms in a light curve automatically and output intermediate predictions

Same as light curve, it can automatically look for the significant peak and fold the light curve to make the class predictions.

Contents:

• Quickstart
  • 1. Load model
  • 2. Load light curve
  • 3. Predict
  • 4. ZTF light curve example:
• Fine-tuning the model
  • Load your own dataset
  • Modify hyperparameters
  • Load baseline model
  • Fine-tune the model
  • Calibration
• API reference
  • deep_lc

Indices and tables

• Index

• Module Index

• Search Page