-------- Original Message --------
Subject: RE: egg shape
Date: Tue, 19 Feb 2008 14:29:52 -0800 (PST)
From: Jonathan Krieger <[EMAIL PROTECTED]>
To: <[email protected]>
Fabio,
If you run an extended eigenshape analysis on your phi data as they are,
the first eigenshape axis represents similarity in your dataset. If that
is, e.g., 93% of the variance explained, you can adjust the variance
explained on the other axes as portion of the remaining 7%. I suspect
the adjusted numbers will be closer to those from your tps analysis.
If this bothers you, you can instead subtract the mean shape from all of
your phi functions before performing the EEA, and you'll get basically
the same results (ES1 will be gone, and the other axes will "move one up
the list").
-Jonathan
Jonathan Krieger
Department of Palaeontology
The Natural History Museum
Cromwell Road, London SW7 5BD
email: [EMAIL PROTECTED]
phone: (44) 0207 942-5681
fax: (44) 0207 942-5546
-----Original Message-----
From: morphmet [mailto:[EMAIL PROTECTED]
Sent: Tuesday, February 19, 2008 8:40 PM
To: morphmet
Subject: egg shape
Misdirected post 6 of 7. -mod
-------- Original Message --------
Subject: egg shape
Date: Mon, 11 Feb 2008 08:44:49 -0500
From: morphmet <[EMAIL PROTECTED]>
To: morphmet <[EMAIL PROTECTED]>
-------- Original Message --------
Subject: egg shape
Date: Fri, 8 Feb 2008 11:44:46 -0800 (PST)
From: Fabio Machado <[EMAIL PROTECTED]>
To: [email protected]
Dear all,
I'm working on egg shape difference in birds and I recently I
tried different analyzes, namely Extended Eigenshape Analysis
(EEA) for open curves and Thin-plate-spline (TPS) using
sliding semilandmarks for the GPA.
Initially I though that as EEA is a contour-based analysis,
it would be a better approach, but it end up giving really
contra-intuitive results. The main goal of the analysis is
egg asymmetry and EEA wasn't reflecting it properly. It was
restricted to the 3rd Eigenshape and reflected only 0.5% of
the variance. I than performed a TPS analysis on the same
dataset, and the results were different. This time asymmetry
was the second Relative Warp and reflected almost 25% of the
variance. The other Eigenshapes and Relative Warps all
reflected similar shape differences but with different shape
partitions.
My mains doubts are: Is this possible, or I am doing some of
the analyzes wrong? If not, is there any particular reason
why the results differ? I can give more details to anyone in
private, if necessary.
cheers,
--
Fábio de Andrade Machado
[EMAIL PROTECTED]
Laboratório de Herpetologia
Museu de Zoologia da USP
Av. Nazaré, 481, Ipiranga
São Paulo, SP, 04263-000
+55 11 61658120
+55 11 82631029
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
