StratifiedMedicine (version 0.2.3)

submod_rpart: Subgroup Identification: CART (rpart)

Description

Uses the CART algorithm (rpart) to identify subgroups. Usable for continuous and binary outcomes. Option to use the observed outcome or PLEs for subgroup identification.

Usage

submod_rpart(Y, A, X, Xtest, mu_train, minbucket = floor(dim(X)[1] *
  0.1), maxdepth = 4, outcome_PLE = FALSE, family = "gaussian", ...)

Arguments

Y

The outcome variable. Must be numeric or survival (ex; Surv(time,cens) )

A

Treatment variable. (a=1,...A)

X

Covariate space.

Xtest

Test set

mu_train

Patient-level estimates (See PLE_models)

minbucket

Minimum number of observations in a tree node. Default = floor( dim(train)[1]*0.05 )

maxdepth

Maximum depth of any node in the tree (default=4)

outcome_PLE

If TRUE, use PLE as outcome (mu_train must contain PLEs). Else use observed outcome Y

family

Outcome type ("gaussian", "binomial"), default is "gaussian"

...

Any additional parameters, not currently passed through.

Value

Trained rpart (CART).

• mod - rpart model as partykit object

References

Breiman L, Friedman JH, Olshen RA, and Stone CJ. (1984) Classification and Regression Trees. Wadsworth

Examples

# NOT RUN {
# }
# NOT RUN {
library(StratifiedMedicine)

## Continuous ##
dat_ctns = generate_subgrp_data(family="gaussian")
Y = dat_ctns$Y
X = dat_ctns$X
A = dat_ctns$A

require(rpart)
res_rpart1 = submod_rpart(Y, A, X, Xtest=X)
res_rpart2 = submod_rpart(Y, A, X, Xtest=X, maxdepth=2, minbucket=100)
plot(res_rpart1$mod)
plot(res_rpart2$mod)
# }
# NOT RUN {

# }