Learn R Programming

PoSIAdjRSquared (version 0.1.0)

construct_selection_event: Construct selection event

Description

This function contructs the selection event by computing c_k, d_k and e_k which are the constants in the quadratic inequalities which characterize the model selection event. The function is used internally by the function solve_selection_event, which returns the intervals of the OLS estimator where the selection event takes place.

Usage

construct_selection_event(a,b,R_M_k,kappa_M_k,R_M_phat,kappa_M_phat)

Value

c_k

Constant c_k in the quadratic inequality c_k*Z^2+d_k*Z+e_k>=0 which characterizes the model selection event of the selected model compared to model k (see Lemma 1 for details)

d_k

Constant d_k in the quadratic inequality c_k*Z^2+d_k*Z+e_k>=0 which characterizes the model selection event of the selected model compared to model k (see Lemma 1 for details)

e_k

Constant e_k in the quadratic inequality c_k*Z^2+d_k*Z+e_k>=0 which characterizes the model selection event of the selected model compared to model k (see Lemma 1 for details)

Arguments

a

Residual vector of type "matrix" and dimension nx1 (see Lemma 1 for details)

b

Vector of type "matrix" and dimension nx1: useful in orthogonal decomposition of y (see Lemma 1 for details)

R_M_k

The orthogonal projection matrix of model k

kappa_M_k

Adjustment factor for model complexity kappa of model k

R_M_phat

The orthogonal projection matrix of the selected model

kappa_M_phat

Adjustment factor for model complexity kappa of the selected model

References

Pirenne, S. and Claeskens, G. (2024). Exact Post-Selection Inference for Adjusted R Squared.