Different results in PCA

52 views
Skip to first unread message

Bruno Ferreto Fiorillo

unread,
Apr 29, 2024, 1:59:26 AM4/29/24
to geomorph R package

Dear everyone, I just started working with GM, and due to my initial lack of confidence in conducting analyses, I opted to run them using different software to validate the results.

Initially, I measured my meshes using Meshlab and Slicer to ensure the accuracy of my scale. The dimensions of the individuals in both software programs closely matched those obtained with a caliper, affirming the correctness of the scale.

Subsequently, I utilized Checkpoint (Stratovan) to pplace my landmarks and exported the coordinates of my specimens at their original scale. This software suppose to export coordinates with the scale already incorporated, as indicated by the company's support team. However, upon conducting the Generalized Procrustes Analysis and PCA in two different softwares, MorphoJ and R (geomorph package), I encountered starkly divergent results.

Could anyone offer insight into what might be causing this disparity? I thank you all in advance, Best regards, Bruno

Adams, Dean [EEOB]

unread,
Apr 29, 2024, 9:39:23 AM4/29/24
to geomorph-...@googlegroups.com

Bruno,

 

Your email is unclear as to what precisely was different. Did you use different landmarking schemes in each analytical package for the PCA? Was it the same data in both packages for the PCA?  At what stage, and in what software, was GPA performed?

 

There are numerous reasons things can differ between MorphoJ and geomorph. Some of these have been mentioned in prior posts (MorphoJ uses full Procrustes fitting, it does not perform sliding of semilandmarks, etc.).


Dean

 

Dr. Dean C. Adams

Distinguished Professor of Evolutionary Biology

Department of Ecology, Evolution, and Organismal Biology

Iowa State University

https://faculty.sites.iastate.edu/dcadams/

phone: 515-294-3834

--
You received this message because you are subscribed to the Google Groups "geomorph R package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to geomorph-r-pack...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/geomorph-r-package/4ef344fe-0080-4f05-a96e-3d3c34d33d49n%40googlegroups.com.

Reply all
Reply to author
Forward
0 new messages