-------- Original Message --------
Subject: Re: Data format in weight matrix from tpsRelW
Date: Wed, 24 Dec 2008 04:47:37 -0800 (PST)
From: Vida Jojic Sipetic <[email protected]>
To: <[email protected]>
References: <[email protected]>
Dear Mark,
when you have opened your Weight matrix in Excel you actually got correct
number of shape variables (2k-4, k-number of landmarks) in your case
2*17-4=30 (first 28 are nonuniform ones, while last two are UNIX and UNIY).
I suppose that your Excel Worksheet looks something like this: first row:
cell A1 (Weight matrix (rows are specimens, cols are x,y pairs), second row:
cell A2 (1) cell B2 (241L) cell C2 (30L) cell D2 (0), third row (IDs for
specimens or values of 0), fourth row (1X, 1Y, 2X, 2Y,....14X, 14Y, UniX,
UniY), fifth row: A5-J5 (1x, 1y, 2x, 2y,...5x, 5y), sixth row: A6-J6 (6x,
6y, 7x, 7y, ...10x, 10y), seventh row: A7-J7 (11x, 11y, 12x, 12y, 13x, 13y,
14x, 14y, UNIX, UNIY). If you want to arrange these variables in a manner
''one row-one specimen'' there are several steps which would make this
faster. First delete first 4 rows, then select a column A and insert one
column. In new inserted A1 cell put a letter A, in A2 put B, in A3 put C.
Then select this block of cells (A1, A2, A3) and double-click on the black
cross in the lower right corner of selected area. You shoud get in a column
A: A, B, C, A, B, C, A, B, C,......A, B, C). Then, select whole Excel
Worksheet and sort by column A. You should get in a column A: A, A,
A,.....B, B, B,....C, C, C. Then, cut block of cells (B242-K482) and paste
it in L1. Also, cut block of cells (B483-K723) and paste it in V1. Delete
column A with A, A,...B, B, C, C. In this way you get 28 Partial Warps plus
2 uniform components (UNIX and UNIY) of your 241 specimens which you can
import in any statistical software and further analyze.
Regarding your second question: There is a paper ''Body shape variation in
cichlid fishes of the Amphilophus citrinellus species complex'' by
Klingenberg et al., Biological Journal of the Linnean Society, 2003, 80,
397-408 (http://www.flywings.org.uk/papers_page.htm)
I hope this help,
Best wishes,
Vida
Vida Jojic Sipetic
Department of Genetic Research
Institute for Biological Research ''Sinisa Stankovic''
Bulevar despota Stefana 142
11060 Belgrade
Serbia
----- Original Message -----
From: "morphmet" <[email protected]>
To: "morphmet" <[email protected]>
Sent: Tuesday, December 23, 2008 8:56 PM
Subject: Data format in weight matrix from tpsRelW
-------- 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
--
Replies will be sent to the list.
For more information visit http://www.morphometrics.org
