jakobrunge/tigramite
Python package for causal discovery and inference on time series data using conditional independence tests and PCMCI methods.
Velocity · 7d
+0.5
★ / day
Trend
→steady
star history
Tigramite provides algorithms for causal discovery from time series datasets, including PCMCI, PCMCIplus, LPCMCI, and RPCMCI methods. It implements various conditional independence tests (ParCorr, GPDC, CMIknn) and a CausalEffects class for estimating causal effects and mediation. The library is used for discovering causal relationships in time series across domains like climate science, neuroscience, and economics.