-------- Original Message --------
Subject: Re: Ideal number of semilandmarks for mixed
landmark-semilandmark analysis
Date: Thu, 6 Mar 2008 09:03:26 -0800 (PST)
From: Fabio Machado <[EMAIL PROTECTED]>
To: [email protected]
References: <[EMAIL PROTECTED]>
Still on that subject... I'm also working with landmark/semilandmarks
data sets and if anyone has any insight on that (appropriate number
of semilandmarks) I would appreciate also.
On a somewhat related subject: IF the appropriate number of
semilandmarks is too great compared to the number of landmarks, and
the curve is dislocated to one side of the landmark configuration,
wouldn't that produce a "Pinocchio effect"? I know that it would only
happen if I have highly dislocated landmarks, but that is what happens
in my case (with the addition of the semilandmarks). If so, can I
correct it?
I thank any comment on this subjects.
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
2008/2/26, morphmet <[EMAIL PROTECTED]>:
-------- Original Message --------
Subject: Ideal number of semilandmarks for mixed landmark-semilandmark
analysis
Date: Tue, 26 Feb 2008 11:52:59 -0800 (PST)
From: thimacek <[EMAIL PROTECTED]>
To: morphmet <[email protected]>
Hello friends,
I'm using a mixed landmark/semilandmark approach to assess morphological
differences of canid scapulae, and I'd like to know if there is any kind
of method to estimate the ideal number of semilandmarks to use, based
primarily on the outline to be analyzed (and maybe on the number of
landmarks - I think many semilandmarks will make the landmarks "less
important", right?)
A co-worker has told me of a method for evaluating the number of
points for extended eigenshape analyses, and we were planning on trying
that same estimation method in this case. Is there a problem with this
approach? Are there better ways to do this?
Thanks a lot for the help!
Thiago
Undergraduate Student at the Laboratory of Geometric Morphometrics - São
Paulo University Zoology Museum, Mammalogy Section.
--
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
