I am new to Geometric Morphometrics, and I'm not sure this is possible or doable.
I have 133 photos with 16 landmarks and 23 semilandmarks of two species and admixture individuals. Their group means differ significantly (done through procD.lm() and pairwise() functions on the superimposed coordinates).
My next step is to calculate not centroid size, but group centroid for each of the two species, so to then calculate the distance between each admixture individuals' point configuration to that group centroid and make an index based on this distance.
Thank you for your time in advance
Hi Joana!
Normally, what it is done is computing the PC score means for each of the groups you are including. You can do that easily in R with the apply() function using the FUN=mean (there are a few tutorials online about this) on the data frame that includes your PC scores. Then, you can use the dist() function to compute the euclidean distance between each individual in a group and the group centroid.
If you are also interested in Mahalanobis distances, you might use the mahalanobis() function of the mvnorm package, but there are several other packages and function that allow you to do the same
Hope it helps
If I understand your question correctly, you want to calculate the Procrustes distance between the shape of each individual in your admixture sample to the mean shape of that group (i.e., the "group centroid" is the "group mean shape".) The function to calculate the mean shape of a group is "mshape" in geomorph. To calculate the Procrustes distance between any two shapes, you can use the procdist function in the shapes package
dist1