oob: Constructs pseudo-outcome, runs regression and causal forests...

In farbmacher/LATEtest: Local Instrument Validity Tests with Causal Forests

Description

Constructs pseudo-outcome, runs regression and causal forests and predicts out-of-bag (oob)

Usage

oob(est, y, huge, tree_fraction, minsize, type)

Arguments

est	estimation sample
y	relevant interval of Y
huge	if set to TRUE, model for orthogonalization learned on random subset of data (size defined in tree_fraction)
tree_fraction	fraction of the data used to build each tree of causal forest (and if huge==T, also used for regression forests); default=0.5
minsize	causal forest insists on at least "minsize" treated and "minsize" control observations per leaf
type	indicates whether we consider D=1 or D=0

Value

pseudo-outcome, oob predictions of pseudo-outcome, oob estimates of causal forest

farbmacher/LATEtest documentation built on Nov. 20, 2020, 10:13 a.m.