i managed to figure it out SE = diag(sqrt(E)) where E is the variance-covariance matrix
Obtaining Heteroscedastic regression coefficients and their properties
16 views (last 30 days)
the picture attaches shows the normal regression line results using the fitlm :estimates of the regression coefficients, the Standard error of the coefficients SE, the tstat and the p value corresponding to the tstat value.
Now im trying to get the same parameters using Heteroscedastic linear regression, ive managed to calculate the estimates of the regression coefficients but im struggling with the errors, any help with any of the parameters will be greatly appreciated