News¶

Updated at: 2026-06-23 17:26:00 CEST

v3.0.1 released¶

June 23, 2026 — v3.0.1 has been released!

Changes:

Add new backends: lfilter (using transfer functions to decouple xi into its elements, then scipy lfilter per element) and cupy (using GPUs & CUDA, if available)
Add generalization to more than 1 dimension (NDCompositeCosts) and corresponding backends
Add new polynomial ALSSMs (Legendre and Meixner)
Add new steady state calculation methods
Add new minimization method
Add rls.fit() and rls.convolve() as composite-functions
Add new examples
Switched docu to mkdocs
Update API of trajectories and windows

November 07, 2022 — v2.1.1 has been released!

What's Changed:

added assertion message for functions minimize_v() by @fwaldmann in https://github.com/lmlib/lmlib/pull/19
added function filter_minimize_v() by @fwaldmann in https://github.com/lmlib/lmlib/pull/18
Documentation fix minimize_h() by @fwaldmann in https://github.com/lmlib/lmlib/pull/17
Down-sample rate option for load_lib_csv, load_lib_csv_mc, load_csv, … by @fwaldmann in https://github.com/lmlib/lmlib/pull/16
Segment(gamma added) by @fwaldmann in https://github.com/lmlib/lmlib/pull/15

Full Changelog: https://github.com/lmlib/lmlib/compare/v2.1.0...v2.1.1

August 15, 2022 — v2.1.0 has been released!

New Examples added TSLM (Two Sided Line Model) Class added Bugfix and print out modifications Documentation

July 04, 2022 — v2.0.2 has been released!

JIT recursions update added examples (jit and window weights)

bugfixes:

June 28, 2022 — v2.0.1 has been released!