[MORPHMET] RE: Testing for sexual dimorphism

[Adams, Dean [EEOBS]]

Miranda,
The required inputs for gpagen do not change based on how you have organized 
the data or how you collected it. From your description, it seems you have 
incorrectly specified the input components.
First, the slide= option is for one to specify a ‘sliders’ matrix, in the style 
used in TPSRelw (a m X 3 matrix specifying which landmarks are semilandmarks 
and in which directions they should be slid). From your email you seem to have 
this in a csv file, so read that into R and use this matrix.
Second, the 3D array of coordinates should contain the xy (or xyz) coordinates 
of both the fixed points as well as the semilandmarks. They do not get input 
separately, but in 1 matrix. In your case, since you seem to have the fixed 
landmarks in one matrix and the semilandmarks in another, you will need to 
concatenate these prior to performing gpagen.
Best,
Dean
Dr. Dean C. Adams
Professor
Department of Ecology, Evolution, and Organismal Biology
   Department of Statistics
Iowa State University
www.public.iastate.edu/~dcadams/
phone: 515-294-3834

From: Karban, Miranda E [mailto:miranda-utzin...@uiowa.edu]
Sent: Wednesday, April 29, 2015 10:41 AM
To: morphmet@morphometrics.org
Subject: [MORPHMET] RE: Testing for sexual dimorphism

Dear Morphometricians,
I am attempting to upload 2D landmark and semi-landmark data into R (geomorph 
package), and I was wondering if anyone could give me advice. In past analyses, 
I have uploaded both a landmark tps file, which contains all of the landmark 
and semi-landmark data combined (measured using tpsDig2), along with a sliders 
csv file (made using tpsUtil), which indicates which points should be treated 
as semi-landmarks. I am hoping to start using the "draw background curves" 
(yellow pencil) tool in tpsDig2, which puts the curve information directly into 
the landmark tps file. Does anyone know whether or not I need to indicate 
within R that these curves exist? And if so, how?

I've tried using the following functon:
> slide <- as.matrix(read.tps("semilandmarks.tps"))

but when I try to use the gpagen function:
> procrustes <- gpagen(landmarks, curves=slide, PrinAxes = FALSE)

the program crashes.

Thank you in advance for any help you can provide.

Best,
Miranda Karban

--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
To unsubscribe from this group and stop receiving emails from it, send an email 
to 
morphmet+unsubscr...@morphometrics.org.

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org

To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.

Re: [MORPHMET] Testing for sexual dimorphism

[Carmelo Fruciano]

Dear Alex,
I imagine most people would consider MAN(C)OVA and/or permutation  
tests for difference in multivariate means as the typical approach.  
Obviously, the choice of methods might also depend on the design and  
research question. There is a chance that with 10 observations per sex  
you might have limited statistical power when comparing within each  
species.

Best,
Carmelo


"Bjarnason, Alexander"  ha scritto:


Dear morphometricians,
I have an interesting GMM dataset based on the cranial morphology of  
adult New World monkeys (platyrrhines) with a wide sampling of  
genera and species. I have approximately ten male and ten females  
for each species, and am interested in testing for size and shape  
dimorphism in each group. My current thinking is to use a Procrustes  
ANOVA to test for significant differences between males and females  
for each taxa, and wanted to enquire whether this was the standard  
approach with geometric morphometric data, or if an alternative  
approach would be more appropriate?

Any advice greatly appreciated,
Best wishes
Alex Bjarnason

--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org

To unsubscribe from this group and stop receiving emails from it,  
send an email to morphmet+unsubscr...@morphometrics.org.







--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org

To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.

[MORPHMET] RE: Testing for sexual dimorphism

[Karban, Miranda E]

Dear Morphometricians,
I am attempting to upload 2D landmark and semi-landmark data into R (geomorph 
package), and I was wondering if anyone could give me advice. In past analyses, 
I have uploaded both a landmark tps file, which contains all of the landmark 
and semi-landmark data combined (measured using tpsDig2), along with a sliders 
csv file (made using tpsUtil), which indicates which points should be treated 
as semi-landmarks. I am hoping to start using the "draw background curves" 
(yellow pencil) tool in tpsDig2, which puts the curve information directly into 
the landmark tps file. Does anyone know whether or not I need to indicate 
within R that these curves exist? And if so, how?

I've tried using the following functon:
> slide <- as.matrix(read.tps("semilandmarks.tps"))

but when I try to use the gpagen function:
> procrustes <- gpagen(landmarks, curves=slide, PrinAxes = FALSE)

the program crashes.

Thank you in advance for any help you can provide.

Best,
Miranda Karban

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org

To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.

[MORPHMET] Testing for sexual dimorphism

[Bjarnason, Alexander]

Dear morphometricians,
I have an interesting GMM dataset based on the cranial morphology of adult New 
World monkeys (platyrrhines) with a wide sampling of genera and species. I have 
approximately ten male and ten females for each species, and am interested in 
testing for size and shape dimorphism in each group. My current thinking is to 
use a Procrustes ANOVA to test for significant differences between males and 
females for each taxa, and wanted to enquire whether this was the standard 
approach with geometric morphometric data, or if an alternative approach would 
be more appropriate?
Any advice greatly appreciated,
Best wishes
Alex Bjarnason

-- 
MORPHMET may be accessed via its webpage at http://www.morphometrics.org

To unsubscribe from this group and stop receiving emails from it, send an email 
to morphmet+unsubscr...@morphometrics.org.