Software
-
(S-4):
[code]
- dncDagger — A C++ program interfaced in R for so-called exact learning of DAG models (Bayesian networks), showing that exact discovery is polynomial for certain sparse DAGs (e.g. matchings).
-
(S-3):
[docs] [code]
- CStrees — A Python package for context-specific graphical models.
-
(S-2):
[docs] [code]
- Benchpress — Benchpress started out from an idea to solve the benchmarking problem for causal discovery and structure learning once and for all. It has grown into a platform that also enables practitioners and researchers to access state-of-the-art open source algorithms for analysis without any coding. It requires minimal installation and scales seamlessly thanks to Snakemake.
-
(S-1):
[code]
- trilearn — A Python package for decomposable graphical models including structure learning and classification.