CI_length_RD: Minimum Half-length Function

In koohyun-kwon/rdadapt:

Minimum Half-length Function

Description

Calculates the minimum half-length, as a function of the modulus value.

Usage

CI_length_RD(b, C, Xt, Xc, mon_ind, sigma_t, sigma_c, alpha)

Arguments

b	point where the inverse modulus is evaluated at.
C	a scalar smoothness parameter.
Xt	n_t by d design matrix for the treated units.
Xc	n_c by d design matrix for the control units.
mon_ind	index number for monotone variables.
sigma_t	standard deviation of the error term for the treated units (either length 1 or n_t).
sigma_c	standard deviation of the error term for the control units (either length 1 or n_c).
alpha	the desired level of non-coverage probability.

Details

In the paper, the minimum half-length is expressed as a function of δ; here, we use the relationship that δ is the inverse modulus value given b. The returned value will be minimized with respect to b later

Value

the minimum half-length corresponding to the modulus value b

Examples

X <- matrix(rnorm(500 * 2), nrow = 500, ncol = 2)
tind <- X[, 1] > 0 & X[, 2] > 0
Xt <- X[tind == 1, ,drop = FALSE]
Xc <- X[tind == 0, ,drop = FALSE]
sigma <- rnorm(500)^2 + 1
sigma_t <- sigma[tind == 1]
sigma_c <- sigma[tind == 0]
CI_length_RD(1.1, 1/2, Xt, Xc, c(1, 2), sigma_t , sigma_c, 0.05)

koohyun-kwon/rdadapt documentation built on May 8, 2022, 8:49 p.m.