Skip to contents

Main function

One call validates the DAG, classifies roles, fits comparison models, and renders the report.

DAGassist()
Generate and/or export report that classifies nodes, compares models, and (optionally) target causal estimands.

Robustness to DAG uncertainty

Check whether conclusions survive uncertain edge directions or plausibly missing arrows.

pdag_robustness()
Diagnose adjustment-set and role robustness to uncertain edge directions
add_edges_robustness()
Diagnose robustness to added ("missing") edges

Diagnostics and building blocks

Standalone helpers used inside DAGassist() that are also useful on their own.

classify_nodes()
Classify DAG nodes
bad_controls_in()
flag bad controls (mediator/collider/desc of Y) among a candidate set
balance_models()
Balance diagnostics across a list of fitted models

Data

Simulated datasets and their generating DAGs, used throughout the documentation and examples.

toy_data
Toy data for demonstrating causal roles
toy_dag
DAG behind toy_data
turnout_data
Simulated data on income and voter turnout
turnout_dag
DAG behind turnout_data

Console printers for DAGassist objects. You rarely call these directly.

print(<DAGassist_addedge_summary>)
Print an edge-addition (exclusion) robustness summary
print(<DAGassist_pdag_summary>)
Print a PDAG robustness summary
print(<DAGassist_report>)
Print method for DAGassist reports
print(<DAGassist_roles>)
Print node classifications (aligned)
print(<DAGassist_validation>)
Minimal, clean printout for validation results with color coding