----- Forwarded message from carlo.mel...@unina.it -----
Date: Wed, 30 Oct 2013 05:41:24 -0400
From: carlo.mel...@unina.it
Reply-To: carlo.mel...@unina.it
Subject: RE: combining linear and geometric measurements
To: morphmet@morphometrics.org
Hi Bryan,
I think Partial Least Square could be a way to go.You can check for
association between shape variables and measurements and look at the
PLS shape space...for those specimens for which you have only
measurements or shape variables you can project them into the PLS space.
Alternatively, use EDMA based on your landmarks coordinates and
perform an analysis based only on measurements.
If your sample include shape and measurements data for the majority of
the specimens, PLS can certainly be an option.
All the best
Carlo
Quoting morphmet_modera...@morphometrics.org:
>
> ----- Forwarded message from "F. James Rohlf"
> <ro...@life.bio.sunysb.edu> -----
>
> Date: Sun, 27 Oct 2013 08:37:18 -0400
> From: "F. James Rohlf" <ro...@life.bio.sunysb.edu>
> Reply-To: ro...@life.bio.sunysb.edu
> Subject: RE: combining linear and geometric measurements
> To: morphmet@morphometrics.org
>
> The problem with performing a PCA on such a mixture of data types is
> that PCA is not invariant to the units of measurement used. Linear
> measurements are not in the same units as the usual shape variables
> for shapes in Kendall's shape space or tangent space. DFA and CVA do
> not have that problem as their results are scale invariant.
> However, they have the problem of needing larger sample sizes than
> is often practical.
>
> ----------------------
> F. James Rohlf, John S. Toll Professor, Stony Brook University
> The much revised 4th editions of Biometry and Statistical Tables are
> now available:
> http://www.whfreeman.com/Catalog/product/biometry-fourthedition-sokal
> http://www.whfreeman.com/Catalog/product/statisticaltables-fourthedition-rohlf
> Please consider the environment before printing this email
>
>> -----Original Message-----
>> From: morphmet_modera...@morphometrics.org
>> [mailto:morphmet_modera...@morphometrics.org]
>> Sent: Sunday, October 27, 2013 4:39 AM
>> To: morphmet@morphometrics.org
>> Subject: combining linear and geometric measurements
>>
>>
>> ----- Forwarded message from Bryan Stephen McLean <mcl...@unm.edu> -----
>>
>> Date: Fri, 25 Oct 2013 17:35:52 -0400
>> From: Bryan Stephen McLean <mcl...@unm.edu>
>> Reply-To: Bryan Stephen McLean <mcl...@unm.edu>
>> Subject: combining linear and geometric measurements
>> To: "morphmet@morphometrics.org" <morphmet@morphometrics.org>
>>
>> Hi everyone,
>>
>> An issue that I’m sure someone can provide insight on…
>>
>> I have a mix of traditional (linear) measurements and geometric morphometric
>> (2D landmark) data from several rodent taxa that I would like to perform
>> PCA/DFA on to hopefully discriminate groups based on ecomorpholoy.
>>
>> What is the best way to combine these 2 types of data?? Can they simply be
>> appended and PCA performed as usual??
>>
>> Thanks much-
>> -Bryan
>>
>> Bryan McLean
>> Department of Biology and
>> Museum of Southwestern Biology
>> MSC03-2020
>> University of New Mexico
>> Albuquerque, NM 87131
>>
>> ----- End forwarded message -----
>
> ----- End forwarded message -----
>
>
>
>
----- End forwarded message -----
