tehtuner0.3.0 package

Fit and Tune Models to Detect Treatment Effect Heterogeneity

get_mnpp.classtree

Get the MNPP for a Classification Tree

get_mnpp.ctree

Get the MNPP for a Conditional Inference Tree

get_mnpp.lasso

Get the MNPP for a Model fit via Lasso

get_mnpp

Get the MNPP for the Step 2 model

get_mnpp.rtree

Get the MNPP for a Regression Tree

get_theta_null

Permute a dataset under the null hypothesis and get the MNPP

get_vt1

Get the appropriate Step 1 estimation function associated with a metho...

get_vt2

Get the appropriate Step 2 estimation function associated with a metho...

permute

Generate a dataset with permuted treatment indicators

print.tunevt

Print an object of class tunevt

test_null_theta_ctree

Test if a Value Gives a Null Conditional Inference Tree

tune_theta

Estimate the penalty parameter for Step 2 of Virtual Twins

tunevt

Fit a tuned Virtual Twins model

validate_alpha0

Check if alpha0 is a valid input to tunevt

validate_p_reps

Check if p_reps is a valid input to tunevt

validate_Trt

Check if Trt is a valid input to tunevt

validate_Y

Check if Y is a valid input to tunevt

vt1_lasso

Estimate the CATE Using the Lasso for Step 1 of Virtual Twins

vt1_mars

Estimate the CATE Using MARS for Step 1 of Virtual Twins

vt1_rf

Estimate the CATE Using a Random Forest for Step 1 of Virtual Twins

vt1_super

Estimate the CATE Using Super Learner for Step 1 of Virtual Twins

vt2_classtree

Estimate the CATE using a classification tree for Step 2

vt2_ctree

Estimate the CATE using a conditional inference tree for Step 2

vt2_lasso

Estimate the CATE using the Lasso for Step 2

vt2_rtree

Estimate the CATE using a regression tree for Step 2

Implements methods to fit Virtual Twins models (Foster et al. (2011) <doi:10.1002/sim.4322>) for identifying subgroups with differential effects in the context of clinical trials while controlling the probability of falsely detecting a differential effect when the conditional average treatment effect is uniform across the study population using parameter selection methods proposed in Wolf et al. (2022) <doi:10.1177/17407745221095855>.

  • Maintainer: Jack Wolf
  • License: GPL (>= 3)
  • Last published: 2023-04-01