draw_proposal_h: Propose active basis

In omkarakatta/ivqr: Instrumental Variables Quantile Regression

Propose active basis

Description

Indices whose observations have small residuals will be more likely to belong to the active basis, while indices whose observations have extreme residuals will be less likely to belong to the active basis.

Usage

draw_proposal_h(
  weights,
  p,
  num_draws,
  label_bool = FALSE,
  label_skip = 5,
  existing_h = c()
)

Arguments

weights	Vector of weights for each observation
p	Dimension of design matrix in IQR-QR (p_X + p_Phi)
num_draws	Number of draws from proposal distribution of active basis
theta	Tuning parameter

Value

indices in the active basis

omkarakatta/ivqr documentation built on Aug. 20, 2022, 11:04 p.m.