TPC_pl_BIC: Variable Selection via Thresholded Partial Correlation

In TPCselect: Variable Selection via Threshold Partial Correlation

Variable Selection via Thresholded Partial Correlation

Description

Use BIC to select the best s and constant over grids.

Usage

TPC_pl_BIC(y, x, u = NULL, s = 0.05, constant = 1, method = "threshold", ...)

Arguments

y	response vector;
x	covariate matrix;
u	non-parametric variable, should be a vector;
s	a value or a vector that used as significance level(s) for partial correlation test. BIC will be used to select the best s.
constant	a value or a vector that used as the tuning constant for partial correlation test. BIC will be used to select the best constant. constant is treated as 1 when method is "simple".
method	the method to be used; default set as method = "threshold"; "simple" is also available.
...	smoothing parameters and functions: kernel, degree, and bandwidth h.

Value

TPC.object a TPC object, which extends the lm object. New attributes are:

• beta - the fitted coefficients

• selected_index - the selected coefficients indices

Examples

#generate partial linear data
samples <- generate_toy_pldata()
y <- samples[[1]]
x <- samples[[2]]
times <- samples[[3]]

#perform variable selection via partial correlation
TPC.fit = TPC_pl_BIC(y,x,times,0.05,c(1,1.5),method="threshold")
TPC.fit$beta

TPCselect documentation built on July 9, 2023, 6:07 p.m.