----- Forwarded message from David Thulman
<dthul...@gmail.com> -----
Date: Wed, 5 Mar 2014 15:03:07
-0500
From: David Thulman <dthul...@gmail.com>
Reply-To:
David Thulman <dthul...@gmail.com>
Subject: Question about DFA in
MorphoJ
To: morphmet@morphometrics.org
I have been using the discriminant function
analysis in MorphoJ and would like to extract the Mahalanobis distances and
Procrustes distances for each specimen. I have exported the Discriminant
Scores and the Cross-validation Scores, but an not sure how to interpret them.
Are these expressed in standard deviations from a consensus shape? After looking
at the specimens, the ones with scores near zero are harder to differentiate
than those with high positive or negative values, but I'm still unsure how I
should describe these values.
Dave Thulman
That intro was rambling, so here are my specific
questions:
1. Is this possible to extract the Mahalanobis and
Procrustes distances for each specimen in MorphoJ?
2. If not, can someone suggest another program
that can generate that data?
3. How should I interpret the discriminant and
cross-validation scores MorphoJ produces?
Thanks for
your help,
Dave Thulman
----- End forwarded message
-----
