Dear Andrea
For the "PC Scores From Other PCA" procedure, you need a dataset (which
contains the items for which you want to compute the scores) and a PCA.
The dialog box has two drop-down menus, one for selecting the dataset
and the other for selecting the PCA. When you select a dataset, the set
of PCAs that is available in the secod drop-down menu changes: only
those PCAs are shown that are consistent with the dataset in the number
of landmarks, dimensionality, object symmetry etc. So PCAs do appear and
disappear in the drop-down menu depending on what dataset you have selected.
Covariance matrices are only involved as far as a covariance matrix is
necessary to run a PCA. The covariance matrices themselves do not appear
in conection with the procedure.
I hope this helps.
Best wishes,
Chris
On 12/12/2016 12:18, andrea cardini wrote:
Today is my morphmet-day.
A question, this time, please.
Is there anyone with experience on the "PC Scores From Other PCA" in
MorphoJ?
I am trying to project shape coordinates on a matrix I imported and
attached to a dataset with mean shapes, but the attached dataset does
not show up in the menu: I can see the parent dataset (before
averaging) with the imported covariance matrix but, as soon as I
select the mean shapes (despite being the data to which the covar.
matrix is attached), the imported covar. matrix vanishes. That means
that I can project the parent dataset (non-averaged) onto the vectors
from the imported matrix but can't do it with the mean shapes (which
is my aim).
I'll do it in NTSYSpc or R, but I'd be curious to understand what I am
getting wrong in MorphoJ.
Thanks in advance for any suggestion.
Cheers
Andrea
--
***********************************
Christian Peter Klingenberg
School of Biological Sciences
University of Manchester
Michael Smith Building
Oxford Road
Manchester M13 9PT
United Kingdom
Web site: http://www.flywings.org.uk
E-mail: c...@manchester.ac.uk
Phone: +44 161 2753899
Skype: chris_klingenberg
***********************************
--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
---
You received this message because you are subscribed to the Google Groups "MORPHMET" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to morphmet+unsubscr...@morphometrics.org.
