I have used TPS to digitize landmarks on fish samples for use in geometric morphometrics. I would also like to undertake a traditional (multivariate) analysis of the data by constructing a truss network of distances that lie between each landmark. Is this possible, and how would I go about doing it?
Dear Brent,
You can obtain interlandmark distances simply by applying the formula of euclidean distances in an excel worksheet. There is a software, called tmorphgen 6c, that do it automatically. You can download it from http://www3.canisius.edu/~sheets/moremorph.html. Probably you can also do it with some R scripts.
I hope this information helps you.
Try this
http://www.morpho-tools.net/measLMs_docs.html
Hope this helps.
Back to the basics, you can also measure interlandmark distances manually using tpsDig software, if you have a scale on the image. You can go to the Image Tools > Measure> Set scale (10mm, for instance), and then Modes > measure mode to take linear measurements. However, I would rather recommend some of the faster procedures mentioned above.
If you can find the right version of WinEDMA (1.0.1 beta May 5, 2003 version) you can simply input an XYZ coordinate file and output a file with all unique linear distances (Utilities tab --> Convert --> Input XYZ file --> Output Linear Distance File). Alternatively, you can calculate interlandmark distances with a simple equation: Distance between landmark 1 (i.e. x1, y1, z1) and landmark 2 (i.e. x2, y2, z2) = √((x1 – x2)^2 + (y1 – y2)^2 + (z1 – z2)^2). If using 2D landmarks, just drop the z-coordinate part of the equation.
If you import your dataset in PAST (I suggest using the older version rather than the new one) you can proceed to make a GPA in this software. Once you do that, under the menu voice "Geomet" there are two options that you can use. One is called "Distance from landmark", where you can decide which landmarks you want to inspect, and the other is "All distances from landmarks", which will give you and output with the distances between all landmarks. It works really well and it is rather easy.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology