Page 35 - THE SLOUGHI REVIEW ISSUE 14

T H E S L O U G H I R E V I E W 3 5

Figure 5: PCA of body measurements of the considered sighthound samples.

Principal Component Analysis (PCA) on body measurements:

The graph represented in Figure 5 shows five groups of variables correlated positively.

The first group includes: Tail length, body length, ear length, and waist circumference.
The second includes: Height at rump, height at withers, forearm circumference and neck
length.
The third includes: Chest circumference, tail circumference and muzzle length.

The fourth includes: Neck circumference, head length and head circumference.
The fifth includes: Waist circumference and muzzle circumference.

All variables are correlated positively; there are no negative correlations between variables or
groups.

*Editor’s note: Principal component analysis (PCA) reduces the number of dimensions in large
datasets to principal components that retain most of the original information. It does this by
transforming potentially correlated variables into a smaller set of variables, called principal
components.