Message-ID: <Pine.SOL.4.44.0304070942240.10983-100000@zektor.gpcc.itd.umich.edu>
Date: 2003-04-07T13:49:36Z
From: Thomas W Blackwell
Subject: Details to factor analysis
In-Reply-To: <004501c2fc5f$9d2e5ff0$fe7aa8c0@tower>
I've been hoping someone else would reply. My only suggestion
is: look up the formula for the quantities you want and calculate
them directly, using matrix or vector arithmetic at the R command
line. That's what most people do.
- tom blackwell - u michigan medical school - ann arbor -
On Sun, 6 Apr 2003, Tino Richter wrote:
> Well, it is no problem to use R or S-Plus for a principle
> component analysis (PCA). The function "princomp" works very
> well with my polychoric correlation matrix, that I use for
> this approach on ordinal variables. But I want to inspect the
> Kaiser-Meyer-Olkin-crieterion with the "measure of sampling
> adequancy (MSA)" and the anti-image-correlation matrix.
> I can't find any possibility to compute this. Please give me
> a hint.
>
> Many Thanks!
> Tino
>