Anton Sherwood wrote:
>
> > I only looked at the image and didn't see anything amiss in the
> > metrics 1 case (I'm completely unable to say what the others
> > should look like anyway...)
>
> The point of the image was that cases 0.9 and 1.1 strongly resemble each
> other while 1.0 looks nothing like them.
>
Yes, but the case 1.0 looks exactly like I'd expect so if there is a
problem, it's likely in the others (but I don't think so, changing
metrics is not a continuous operation...)
--
* Doctor Jekyll had something * mailto:ber### [at] iname com
* to Hyde... * http://www.enst.fr/~jberger
*******************************
Post a reply to this message
|