-------- Original Message --------
Subject: Data format in weight matrix from tpsRelW
Date: Tue, 23 Dec 2008 11:18:45 -0800 (PST)
From: Mark H Schmitz <[email protected]>
To: [email protected]
Hello,
I'm attempting to analyze shape differences between two color forms of a
fish species based on 17 landmarks digitized on 241 specimens. I have
samples from two different populations (lakes) of these color forms and
can also subdivide each color form by sex. I used tpsRelw to generate
and save a weight matrix containing the partial warps and uniform
component shape data. The "report" generated from tpsRelw indicates
there are 30 relative warps.
When I attempt to transform the nts weight matrix data into excel (using
Delimited: tab and space) only 5 x,y data pairs appear in columns.
Also, the number of rows representing specimins is equal to 723 which is
3x my number of specimens. Should I not expect to have one row per
specimen with 30 x,y data pairs (representing the 30 relative warps)
plus one x,y pair representing the uniform component? Is there a data
wrapping function that would explain this format?
I have used tpsUtil to "build a variable matrix" and I hope to combine
my variables (lake, color, sex) with the weight matrix and use tpsRegr
to examine the effect of these variables. Does this sound like the best
approach for this type of data? Any advice would be greatly
appreciated! Thanks!
Mark H. Schmitz
PhD Candidate
UW-Milwaukee
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
