tlse: Thresholding Least Squares estimator

In causalTLSE: Thresholding Least Squares Inference for Causal Effects

Thresholding Least Squares estimator

Description

This is the main function to estimate the causal effects using the semiparametric thresholding least squares method.

Usage

tlse(model, selType=c("SLSE","BTLSE","FTLSE"),
     selCrit = c("ASY", "AIC", "BIC"),
     causal = c("ALL","ACT","ACE","ACN"),
     seType=c("analytical", "lm"),
     minPV = function(p) 1/log(p), vcov.=NULL, ...)

Arguments

model	A model of class tlseModel created by the getModel function.
selType	The method for selecting the piecewise polynomial knots.
selCrit	The criterion to select the piecewise polynomial knots.
causal	What causal effect should we compute.
seType	Should the causal effect standard errors be computed based on an analytical expression or based on least squares asymptotics.
minPV	A function to determine the threshold for the significance of the coefficients. It has to be a function of one parameter, which is the average number of knots in the model.
vcov.	An alternative function to compute the covariance matrix of the least squares estimators. The default is the vcov method for lm objects.
...	Additional arguments to pass to vcov.

Examples

data(simData)
mod1 <- setModel(Y~Z|~X1*X2, data=simData)
fit <- tlse(mod1, selType="BTLSE", vcov.=sandwich::vcovHC, type="HC3")

causalTLSE documentation built on March 31, 2023, 3:07 p.m.