| 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 |