QNPy-Latte
Latent ATTEntive Neural Processes for Quasar Light Curves with parametric recovery.
QNPy-Latte is an open-source Python package developed by the SER-SAG-S1 team as part of the Serbian in-kind software contribution to the Vera C. Rubin Observatory / LSST. It models quasar light curves with Attentive Latent Neural Processes (AttnLNPs), clusters them with Self-Organizing Maps (SOMs), and recovers physical parameters of the source - in particular the transfer function - from the model’s latent representation.
It is the successor to QNPy (Conditional Neural Processes), adding a latent path, attention, optional RNN time-encoding, and parameter/transfer-function recovery.
Note
This documentation site is maintained for the SER-SAG-S1 in-kind closeout. The package
source lives at https://github.com/rajuaman1/QNPy_Latte and is installable from PyPI
(pip install QNPy_Latte).
Getting started
User guide
Tutorials
Background
API reference