CausalIQ Analysis API Reference¶
This is the entry point for the API documentation. Each module is documented separately below.
Modules¶
Merge¶
Functions for merging multiple graphs (DAG, PDAG, PDG) into a single Probabilistic Dependency Graph (PDG) with weighted edge probabilities.
Metrics¶
Functions for analysing and comparing causal graphs, including structural metrics, Kullback-Leibler divergence calculations, and Bayesys compatibility metrics.
Graph¶
Enumerations for describing changes made to causal graphs during structure learning, including action types and detailed information recorded with each trace entry.
Trace¶
Detailed tracing of structure learning processes, enabling researchers to record, analyse, and compare the step-by-step evolution of causal graphs during algorithm execution.
Plot¶
Functions for drawing charts using matplotlib and Seaborn from a
summarise CSV output, migrated from the legacy experiments/plot.py
module. The code is split into the plot/ package (property parsing,
axis styling, chart builders and the run_plot entry point), with the
public names re-exported from causaliq_analysis.plot.
CLI¶
Command-line interface for causaliq-analysis operations. A click group in
causaliq_analysis.cli registers each command from
causaliq_analysis.cli.commands, so the individual commands are documented
on their own pages.
Workflow Action¶
Workflow action interface for causaliq-workflow: a thin provider that
validates parameters and delegates to per-action classes registered in
ACTION_CLASSES (migrate_trace, merge_graphs, evaluate_graph, best_graph,
summarise, plot).