Alexander Sirotkin schrieb:
>
> Thanks a lot, it seems to be correct, however...
> load^-1 produces lots of infinite values.
>
> summary(c((p$loadings)^-1))
> Min. 1st Qu. Median Mean 3rd Qu. Max.
> -3.094e+06 -3.747e+01 -6.409e+00 Inf 3.761e+01 Inf
>
> Any idea why ?
Perhaps R=load*load' has low rank? (load' = transposed(load))
Then you still can use the formula with a slight correction
> > load*fsc = zdat // if zdat is unknown
> > fsc = load^-1 * zdat // if fsc is unknown
If load has more rows than columns:
fsc = (load'*load)^-1*load'*zdat
For the opposite problem I don't have a quick solution at hand
Gottfried Helms
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
