-------- Original Message --------
Subject: Re:urgent:separating the shape from the size
Date: Mon, 12 May 2008 17:29:15 -0700 (PDT)
From: thimacek <[EMAIL PROTECTED]>
To: morphmet <[email protected]>
Ali,
Actually, the results of the principal component analysis or the
relative warps analysis (which is similar, though not identical) are
ideally not what you should use to test for relations between size and
shape, since they are already "processed" results, showing you the shape
changes responsible for most of the variation present in your dataset.
Ideally, I'd say you should use the Partial Warps as the dependent
variables for a regression with size as the predictor variable, since
the Partial Warps are the "raw" shape variables (not orthogonal, not
organised in order of "total variation explained", etc).
Unfortunately, I cannot help you with softwares for this analysis, since
I've never worked with 3D data. I'm sure others on the list could help
you with that, however.
Best regards,
Thiago
>
>
> -------- Original Message --------
> Subject: Re:urgent:separating the shape from the size
> Date: Sun, 11 May 2008 04:41:19 -0700 (PDT)
> From: ali mahmood
> Reply-To: [EMAIL PROTECTED]
> To: [email protected]
>
>
>
> Dear Thiago
>
>
>
> Thank you indeed for your kind information, do you mean that I need to
> do the Principal component analysis and then use the engine values as a
> determinants for the shape?, my data are 3D do you have any other
> suggested software? I am waiting for your helpful answer,thank you
again.
>
>
>
> Ali
>
> --- On *Sat, 5/10/08, morphmet //*
> wrote:
>
> From: morphmet
> Subject: Re:urgent:separating the shape from the size
> To: "morphmet"
> Date: Saturday, May 10, 2008, 9:37 PM
>
> -------- Original Message --------
> Subject: Re:urgent:separating the shape from the size
> Date: Sat, 10 May 2008 07:09:37 -0700 (PDT)
> From: thimacek
> To: morphmet
>
>
>
> Ali,
>
> By using geometric morphometrics, you are already separating "size"
> and
> "shape" most, if not all of the time. If you used Landmark methods
> (digitization followed by GPA), you'll have come to a number of
> 'Partial
> Warps' and possibly 'Relative Warps', in addition to the Centroid
> Size.
> The centroid size is a measure of size, while the Partial Warps are the
> variables containing the Shape information. It should be noted that all
> Partial Warps should be analyzed together, as each one of them alone
> normally have no biological meaning per se.
> The Relative warps are a sort of PCA (principal component analysis)
> done
> on the Partial Warps, and as such one may be able to draw conclusions
> by analyzing only the first among them, which represent larger amounts
> of the sample's total variation.
>
> If you want to look for correlation between size and shape - i.e.
> allometry - you may make a multivariate regression using Centroid Size
> (the size variable) as the independent variable and the Partial Warps
> (all of them, which together represent shape) as dependent
> variables. If
> you're making your analysis in 2D, you can do this using the TPSRegr
> program by Rohlf.
>
> I hope this helps!
>
> Thiago
>
> >
> >
> > -------- Original Message --------
> > Subject: urgent:separating the shape from the size
> > Date: Sat, 10 May 2008 04:53:11 -0700 (PDT)
> > From: ali mahmood
> > Reply-To: [EMAIL PROTECTED]
> > To: [email protected]
> >
> >
> >
> > Dear Friends
> >
> >
> >
> > Thank you for your kind help, I send before a question
> regarding the
> > correlation between the the *size* of face and dental arch in
> geometric
> > morphometric means ,and I received the answer from some Friends
> as the
> > told me that I need to use the *centroid size* for both of them to
> > calculate the correlation for the *size * ,but no one answer my
> question
> > regarding the correlation between the *shape *of the face and
> dental
> > arch (because I would like to separate the* size* from the
> *shape, *so
> > what I need to calculate for the_ face_ and then to the_ dental
> arch_ to
> > be used as variables in the correlation for the *shape*?
> >
> >
> >
> >
> >
> > Ali R Al-khatib
> >
> >
> >
> ------------------------------------------------------------------------
> > Be a better friend, newshound, and know-it-all with Yahoo!
> Mobile. Try
> > it now.
> > <
> evt="51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ";
> us.rd.yahoo.com>
> >
> > >
> >
> > --
> > 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
>
>
> ------------------------------------------------------------------------
> Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try
> it now.
> <
evt="51733/*http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ";
us.rd.yahoo.com>
>
> >
>
> --
> 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
