var_i: var_i

In fineiskid/scdlintrend: Effect Size Estimation in SCD Studies with Linear Trends

Description

calculate the variance about the regression line for case i

Usage

var_i(X, y, Sigma, k)

Arguments

X	the (4k, N_i)-dimensional design matrix for case i.
y	the vector of N_i outcomes for case i
Sigma	the AR(1) covariance matrix
k	number of complete sub-phase cycles, as in (AB)^k

Value

sigma_sq_i estimate, A_i an (N_i, N_i)-dimensional matrix (latter value necessary for combining treatment effects across individuals)

Examples

#assuming you've run >sim <- simulate_ABk(...)
sigma_calcs <- var_i(design_1, y_1, sigma_1, k = 2)

fineiskid/scdlintrend documentation built on May 16, 2019, 12:53 p.m.