tboost: T-learner, implemented via xgboost (gradient boosting)

In xnie/rlearner: R-learner for Heterogeneous Treatment Effect Estimation

Description

T-learner learns the treated and control expected outcome respectively by fitting two separate models.

Usage

tboost(
  x,
  w,
  y,
  k_folds_mu1 = NULL,
  k_folds_mu0 = NULL,
  ntrees_max = 1000,
  num_search_rounds = 10,
  print_every_n = 100,
  early_stopping_rounds = 10,
  nthread = NULL,
  verbose = FALSE
)

Arguments

x	the input features
w	the treatment variable (0 or 1)
y	the observed response (real valued)
k_folds_mu1	number of folds for cross validation for the treated
k_folds_mu0	number of folds for cross validation for the control
ntrees_max	the maximum number of trees to grow for xgboost
num_search_rounds	the number of random sampling of hyperparameter combinations for cross validating on xgboost trees
print_every_n	the number of iterations (in each iteration, a tree is grown) by which the code prints out information
early_stopping_rounds	the number of rounds the test error stops decreasing by which the cross validation in finding the optimal number of trees stops
nthread	the number of threads to use. The default is NULL, which uses all available threads
verbose	boolean; whether to print statistic

Examples

## Not run: 
n = 100; p = 10

x = matrix(rnorm(n*p), n, p)
w = rbinom(n, 1, 0.5)
y = pmax(x[,1], 0) * w + x[,2] + pmin(x[,3], 0) + rnorm(n)

tboost_fit = tboost(x, w, y)
tboost_est = predict(tboost_fit, x)

## End(Not run)

xnie/rlearner documentation built on April 11, 2021, 12:49 a.m.