tuning_parameter_selection: Select the tuning parameters

In BIG-S2/GFPLVCM: Genealized functional partinal linear varying-coefficient models

Description

Select the tuning parameters via the roughly estimated mean square errors and do estimation, prediction as well as hypothesis testing with the selected tuning parameters.

Usage

tuning_parameter_selection( lambda_range, h_range, y, x, z, order_1, order_2, breaks, pois_par, grid)

Arguments

lambda_range	a vector of the tuning parameter lambda.
h_range	a vector of the bandwidth.
y	a list of the response with the elements y_ID, y_time_point and y_value.
x	a list of the functional covariate with the elements x_ID, x_time_point and x_value.
z	a list of the scalar covariate with the elements z_ID, z_time_point and z_value.
order_1	order of the B-splines in the s direction.
order_2	order of the B-splines in the u direction.
breaks	knots of the B-splines.
pois_par	intensity of the Possion distribution when generating the time points of the functional parameter.
grid	observation points in the u direction.

Value

final_lambda	The seleted roughness penality tuning parameter.
final_bd	The selected bandwidth.

References

see the paper "Generalized functional partial varying-coefficient model".

BIG-S2/GFPLVCM documentation built on May 23, 2019, 5:01 a.m.