[Gretl-users] GUI points

Sven Schreiber svetosch at gmx.net
Sat Nov 12 17:25:14 EST 2016


Am 12.11.2016 um 23:12 schrieb Allin Cottrell:

> Hmm, I guess so. Maybe worth noting that Belsley's condition number
> calculation involves scaling but not centering. So try the following:
>
> <hansl>
> open hall.gdt
> series c10 = consrat * 10
> series cc = consrat - mean(consrat)
> eval cnumber({const, ewr, consrat})
> eval cnumber({const, ewr, c10})
> eval cnumber({const, ewr, cc})
> </hansl>
>
> Multiplying consrat by 10 makes no difference to the condition number,
> as you say. But centering it reduces the value a great deal. (With
> correlation, of course, neither scaling not centering makes any
> difference.)

(And centering ewr, too, then leads to cnumber values below 2, BTW.)

I haven't looked (or wasn't forced to look) at these concepts in the 
past, but this behavior of the condition number I find odd. It feels as 
if it makes only sense if the data are demeaned. And including a 
constant term in the variable set for the calculation of cnumber 
obviously doesn't do the same thing.

thanks,
sven


More information about the Gretl-users mailing list