API Reference

Import causalcompass as:

import causalcompass

CausalCompass provides data generation functions for both VAR (linear) and Lorenz-96 (nonlinear) models across multiple assumption-violation scenarios. All data generation functions are located under causalcompass.datasets. All algorithms are located under causalcompass.algorithms.

Data Generation

`datasets.vanilla.simulate_var`	Generate time series data from a Vector Autoregressive (VAR) model.
`datasets.vanilla.simulate_lorenz_96`	Generate time series data from the Lorenz-96 dynamical system.
`datasets.measurement_error.simulate_var_with_measure_error`	Generate VAR data with measurement error proportional to data variance.
`datasets.measurement_error.simulate_lorenz_with_measure_error`	Generate Lorenz-96 data with measurement error.
`datasets.mixed_data.generate_mixed_var`	Generate VAR data where a proportion of variables are discretized.
`datasets.mixed_data.generate_mixed_lorenz_96`	Generate Lorenz-96 data where a proportion of variables are discretized.
`datasets.standardized.generate_standardized_var`	Generate VAR data with z-score or min-max normalization applied.
`datasets.standardized.generate_standardized_lorenz_96`	Generate Lorenz-96 data with normalization applied.
`datasets.confounder.simulate_var_with_confounders`	Generate VAR data with cross-lag hidden confounders that introduce spurious correlations.
`datasets.confounder.simulate_lorenz_with_confounders`	Generate Lorenz-96 data with hidden confounders.
`datasets.missing.generate_missing_var`	Generate VAR data with missing values and interpolation.
`datasets.missing.generate_missing_lorenz_96`	Generate Lorenz-96 data with missing values and interpolation.
`datasets.trendseason.simulate_var_with_trend_season`	Generate VAR data with additive trend and seasonal components.
`datasets.trendseason.simulate_lorenz_with_trend_season`	Generate Lorenz-96 data with additive trend and seasonal components.
`datasets.nonstationary.simulate_nonstationary_var`	Generate VAR data with time-varying noise variance (driven by Gaussian Process).
`datasets.nonstationary.simulate_nonstationary_var_timevarying_coef`	Generate VAR data with both time-varying coefficients and time-varying noise.
`datasets.nonstationary.simulate_nonstationary_lorenz_96`	Generate Lorenz-96 data with time-varying noise variance.

Algorithms

`algorithms.VAR`	Granger causality-based causal discovery method that fits a Vector Autoregressive model and infers causal relationships from the estimated coefficient matrix.
`algorithms.LGC`	Granger causality-based causal discovery method that augments the VAR model with a Lasso penalty term.
`algorithms.VARLiNGAM`	Noise-based causal discovery method that combines Vector Autoregressive model with Linear Non-Gaussian Acyclic Model (LiNGAM) for time series data.
`algorithms.PCMCI`	Constraint-based causal discovery method that combines the PC algorithm with Momentary Conditional Independence tests for time series data.
`algorithms.DyNotears`	Score-based causal discovery method that extends the NOTEARS framework to dynamic (time series) settings using continuous optimization with acyclicity constraints.
`algorithms.NTSNotears`	NTS-NOTEARS is a score-based nonlinear extension of DYNOTEARS.
`algorithms.TSCI`	Topology-based causal discovery method that leverages Takens’ state-space reconstruction theory to infer causality.
`algorithms.CMLP`	Deep learning-based causal discovery method that uses component-wise MLPs to model nonlinear Granger causality.
`algorithms.CLSTM`	Deep learning-based causal discovery method that uses component-wise LSTMs to model nonlinear Granger causality.
`algorithms.CUTS`	CUTS is a deep learning-based causal discovery method tailored for irregular time series data.
`algorithms.CUTSPlus`	CUTS+ extends the original CUTS framework.