betaTildeSq: Bias Correct OLS model with squared variable

In georgieevans/PrivacyUnbiased: Unbiased Estimates from Privacy Protected Data

Description

This function takes OLS coefficients from a model including a squared variable as inputs and bias corrects according to the noise matrix, S

Usage

betaTildeSq(coef, S, X, N, index_sq)

Arguments

coef	Vector of OLS coefficients
S	Matrix of error terms. Dimensions must be length(coef) x length(coef)
X	Design matrix that produced OLS coefficients
N	Number of rows in X
index_sq	Column index of X covariate that is squared

Value

Returns a list:

bias_correct	Vector of bias corrected coefficients
x_prime_x	X'X matrix to be used in variance estimation

georgieevans/PrivacyUnbiased documentation built on May 22, 2021, 1:01 p.m.