Policy Learning Under Constraint

---

Documentation for package ‘PLUCR’ version 0.1.0

• DESCRIPTION file.
• User guides, package vignettes and other documentation.

Help Pages

binary_S_p	Constraint function for binary policy
check_data	Check input data for validity
CVFolds	CVFolds (from SuperLearner package)
delta_mu_constant	Constant Conditional Average Treatment Effect estimator for Y
delta_mu_linear	Linear-shaped Conditional Average Treatment Effect estimator for Y
delta_mu_mix	Mixed-shape Conditional Average Treatment Effect estimator for Y
delta_mu_null	Null Conditional Average Treatment Effect estimator for Y
delta_mu_realistic	Realistic Conditional Average Treatment Effect estimator for Y
delta_mu_threshold	Thresholded-shaped Conditional Average Treatment Effect estimator for Y
delta_nu_linear	Linear-shaped Conditional Average Treatment Effect estimator for Xi
delta_nu_mix	Mixed-shaped Conditional Average Treatment Effect estimator for Xi
delta_nu_realistic	Realistic Conditional Average Treatment Effect estimator for Xi
delta_nu_satisfied	Computes the difference in expected outcomes under treatment and control.
delta_nu_threshold	Thresholded Conditional Average Treatment Effect estimator for Xi
estimate_mu	Estimate mu
estimate_nu	Estimate nu
estimate_ps	Estimate propensity score
estimate_real_valued_mu	Estimate real-valued mu
FW	Frank-Wolfe algorithm
generate_data	Synthetic data generator and functions generator
generate_realistic_data	Realistic synthetic data generator and functions generator
get_opt_beta_lambda	Select Optimal Beta and Lambda Combination
grad_Lagrangian_p	Gradient of the objective function
grad_Lagrangian_p_X	Gradient of the objective function
HX	Compute the Inverse Propensity Score Weight (IPW)
Lagrangian_p	Objective function taking the form of a Lagrangian
learn_threshold	Learn Optimal Decision Threshold
lwr_upper_bound_estimators	Lower and upper bound estimators for policy value and constraint
main_algorithm	Main algorithm
make_psi	Generate psi function
model_Xi_linear	Linear treatment effect on Xi Component Function
model_Xi_mix	Mixed treatment effect on Xi component function
model_Xi_realistic	Realistic treatment effect on Xi Component Function
model_Xi_satisfied	Low treatment effect on Xi
model_Xi_threshold	Thresholded treatment effect on Xi component function
model_Y_constant	Constant treatment effect on Y component function
model_Y_linear	Linear treatment effect on Y component function
model_Y_mix	Mixed treatment effect on Y component function
model_Y_null	No treatment effect on Y component function
model_Y_realistic	Realistic treatment effect on Y component function
model_Y_threshold	Thresholded treatment effect on Y component function
naive_approach_algorithm	Naive approach main algorithm
Optimization_Estimation	Iterative optimization procedure
oracular_approach_algorithm	Oracular approach main algorithm
oracular_process_results	Oracular evaluation of a policy
phi	Normalize a Matrix by Column Min-Max Scaling
phi_inv	Inverse Min-Max Normalization
plot_metric_comparison	Plot metric values for comparison
plot_realistic	Plot realistic data setting
predict.SL.grf	predict.SL.grf
process_results	Evaluate a policy
R_p	Risk function for Conditional Average Treatment Effect (CATE)
SGD	Stochastic Gradient Descent (SGD) algorithm
sigma_beta	Link function
sigma_beta_prime	Derivative of link function
SL.grf	SL.grf
SuperLearner.CV.control	SuperLearner.CV.control (from SuperLearner package)
synthetic_data_plot	Plot synthetic data setting
S_p	Constraint function
update_mu	Update mu via augmented covariate adjustment
update_mu_XA	Update mu via augmented covariate adjustment for fixed X
update_nu	Update nu via augmented covariate adjustment
update_nu_XA	Update nu via augmented covariate adjustment for fixed X
visual_treatment_plot	Visualize treatment assignment probability
V_p	Oracular approximation of value function
V_Pn	Estimation of policy value