Hello every one,
I wanted to let you know a new program has been posted for the estimation of the leave one out error, and estimation of CV statistics for linear models.
In specific, it provides the same statistics as the user written program -loocv-, but only for after the -regress- command, and gives you the option to estimate the leave one out predicted error.
It does this using the shortcut based on the leverage statistics, for which you may find it much faster than loocv in large samples.
Hopefully you find it useful.
Best regards
Fernando
I wanted to let you know a new program has been posted for the estimation of the leave one out error, and estimation of CV statistics for linear models.
In specific, it provides the same statistics as the user written program -loocv-, but only for after the -regress- command, and gives you the option to estimate the leave one out predicted error.
It does this using the shortcut based on the leverage statistics, for which you may find it much faster than loocv in large samples.
Hopefully you find it useful.
Best regards
Fernando
