Library: | eiv |
See also: | eivknownvaru |
Macro: | eivknownvarumod | |
Description: | eivknownvarumod presents modified moment estimates of parameters for the measurement error models, which has only one variable x. The variance of measurement error sigma_u is known. All of the variables obey normal distributions.The theoretical discussion see section 2.5 of Fuller (1987) page 163-172. |
Usage: | {mux,beta1,beta0,sigmax, sigmae} =eivknownvarumod(alpha,X,Y,sigmau) | |
Input: | ||
alpha | scalar | |
X | n x 1 matrix, the design variables | |
Y | n x 1 matrix, the response | |
sigmau | scalar, the variance of measurement error sigma_u | |
Output: | ||
mux | scalar, the mean value of X | |
beta1 | scalar, the estimate | |
beta0 | scalar, the estimate | |
sigmax | scalar, the estimate of the variance of X | |
sigmae | scalar, the estimate of the variance of error e |
library("eiv") n = 100 randomize(n) x= normal(n)*9 w=x+9*normal(n) y =0.9+0.8*x+0.01*normal(n) alpha=5 sigmau=81 gest=eivknownvarumod(alpha,w,y,sigmau) gest.mux gest.beta1 gest.beta0 gest.sigmax gest.sigmae
gest.mux=-0.93396; gest.beta1=0.83504 gest.beta0=0.36572 gest.sigmax=64.17; gest.sigmae=0.36572
Library: | eiv |
See also: | eivknownvaru |