missing value handling & computational precision

From: Ian Martin (idmartin@golden.net)
Date: Wed Mar 30 2005 - 18:30:06 MEST

Hello/Bonjour -

I have a question about ADE's handling of missing values in correlation
PCA. Are missing values deleted pairwise in computation of the
correlation matrix, or are means substituted? What is the default and
is there facility to change this? I haven't been able to find this
answer in the documentation pdfs, but perhaps I've not been looking in
the correct place.

Also, what is the computational precision of ADE-4 -- double precision
(64-bit, real*8) or long precision (128-bit, real*16)?

I'm trying to compare some PCA results to those obtained from SPSS.


Ian D. Martin, Ph.D.
Aquatic Ecologist

This archive was generated by hypermail 2b30 : Tue Oct 18 2005 - 10:35:39 MEST