Library: | eiv |
See also: | eivknownatt eivknownvaru |
Macro: | eivknownratue | |
Description: | eivknownratue presents the moment estimates of the parameters in the measurement error models, which has only one variable x. The ratio of two variances of the two measurement errors is known. All of the variables obey normal distributions. See Fuller(1987), section 1.3. |
Usage: | {mux,beta1,beta0,sigmax,sigmau,sigmae} = eivknownratue(X,Y,delta) | |
Input: | ||
X | n x 1 matrix, the design variables | |
Y | n x 1 matrix, the response | |
delta | scalar, the ratio of the variance of two errors | |
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 | |
sigmau | scalar, the estimate of the variance of u | |
sigmae | scalar, the estimate of the variance of 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) delta =0.0001/81 ; The ratio of var(u)/var(e) gest=eivknownratue(w,Y,delta) gest.mux gest.beta1 gest.beta0 gest.sigmax gest.sigmau gest.sigmae
gest.mux=-0.93396; gest.beta1=0.92972 gest.beta0=0.45415;gest.sigmax=61.31 gest.sigmau=83.86; gest.sigmae=0.00010353
Library: | eiv |
See also: | eivknownatt eivknownvaru |