CLI: Uplift

causal-toolkit uplift --data DATA.csv --treatment T --outcome Y --covariates X1 X2 [OPTIONS]

Options

Option	Short	Required	Description
--data	-d	Yes	Input CSV
--treatment	-t	Yes	Treatment column
--outcome	-y	Yes	Outcome column
--covariates	-c	Yes	Covariate columns (space-separated)
--model	-m	No	causal_forest, two_model, dr_learner, class_transformation (default: causal_forest)
--plot	-p	No	qini, gain (default: qini)
--out	-o	No	Output directory (default: ./uplift)

Examples

# Causal Forest (recommended)
causal-toolkit uplift -d data.csv -t treatment -y outcome -c age income education \
    -m causal_forest

# DR-Learner
causal-toolkit uplift -d data.csv -t treatment -y outcome -c age income education \
    -m dr_learner --plot gain

# With custom params (via config)
causal-toolkit uplift -d data.csv -t treatment -y outcome -c age income education \
    -m causal_forest --n_estimators 1000 --min_samples_leaf 20

Output

Uplift Metrics: {'qini': 0.342, 'auuc': 0.187, 'gain_at_10pct': 2.45, 'gain_at_50pct': 1.12}

Files: - uplift/uplift_metrics.json - uplift/qini_curve.png or gain_curve.png