lmlib Documentation

Version: v3.0.1 — Open-Source and for Python.

Official documentation for lmlib, the signal processing library using linear state space models for computationally efficient event detection, signal convolution or correlation, and signal decomposition tasks in time series (single- and multi-channel) and N-dimensional data (including images, videos, ...).

This library supports scipy, JIT (Just-in-Time) compilation and GPU accelerators.

Detailed information and scientific background are available in our publications (see Bibliography).

Typical Fields of Applications:

• Recursive least squares filters (options: weighted, windowed)
• Recursive linear filters (convolutions and correlations)
• Shape Detection, event and feature detection, likelihood testing
• for one-dimensional and multi-dimensional data sets