Ioannis Maniakas <biojohn...@gmail.com> ha scritto:
Hello everyone. As I am new in the field of the geometric
morphometrics, I am expressing my naive problem..
I am trying to analyze differences in the shape of two simple
separate curves. These describe the radii of circles fitting to the
condylar margins on the proximal and distal articular surfaces of
the medial astragalar side respectively.
I think I have followed the standard procedure. The recorded curves
were resampled (always from a common homologous starting and ending
position), granting the same number of evenly spaced 50 points (“by
length” option in tpsDig) across all specimens.
I would double-check that corresponding points across different
specimens sample the same structure.
Afterwards, the operation of converting into landmarks was conducted
with the employment of the tpsUtil program resulting in 50 landmark
configurations.
So, my question is this: can I import their coordinates into Morphoj
or even Past and ran the whole analysis there, including the
superimposition? Or do I have to treat these data in an other way?
No. If these are, in fact, semilandmarks, these should be treated
accordingly (sliding).
I have tried to make a slider file for both curves (2 landmarks at
the two ends of the curves and 48 sliding landmarks between). As I
load these nts files with the tps file with my 50 lanmarks in
tpsrelw and perform the analysis there, the consensus and the rest
results are wrong. More specifically, the position of landmarks has
changed dramatically, as some of them have overslided and there are
not placed equidistantly.
I would check different criteria for sliding (menu Options of tpsRelW).
Also, there are various discussions of criteria and relative
advantages you might want to consult, such as
Perez et al 2006 - Journal of Anatomy
Gunz & Mitteroecker 2013 - Hystrix
To memory, the latter explicitly discusses a case which sounds similar
to yours.
Also, it seems that right and left specimens are not getting aligned at all.
I guess I would align them separately (unless you want to reflect and
relabel one of the two sides prior to superimposition, which I imagine
you'd have to do "manually"/scripting).
This was not the case, when I simply treated them as typical
landmarks in morphoj..
That is because the Procrustes fit in MorphoJ includes reflection. You
can find a description of this in the MorphoJ user guide.
I hope this helps,
Carmelo
--
Carmelo Fruciano
Postdoctoral Fellow - Queensland University of Technology - Brisbane,
Australia
Honorary Fellow - University of Catania - Catania, Italy
e-mail c.fruci...@unict.it
http://www.fruciano.it/research/
--
MORPHMET may be accessed via its webpage at http://www.morphometrics.org
---
You received this message because you are subscribed to the Google Groups "MORPHMET" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to morphmet+unsubscr...@morphometrics.org.
