PCA graph of Eurogenes K15 by Orange

[21993]

https://gen3553.pagesperso-orange.fr/ADN/K15V4.png
My question is, how could the producer of this graph place the principle components (such as Atlantic, Amerindian, West Asian, etc) onto the chart? I mean, those principle components seemed arbitrarily located onto the chart because as far as I know, when doing principle component analysis, principle components should be equidistant from their adjacent components and therefore form a circle in a 2 dimensional plane. Do you also think the principle components are arbitrarily located on the PCA graph of Eurogenes K15 by Orange?

You can reach the full website here: https://gen3553.pagesperso-orange.fr/ADN/K15.htm

[vbnetkhio]

https://gen3553.pagesperso-orange.fr/ADN/K15V4.png
My question is, how could the producer of this graph place the principle components (such as Atlantic, Amerindian, West Asian, etc) onto the chart? I mean, those principle components seemed arbitrarily located onto the chart because as far as I know, when doing principle component analysis, principle components should be equidistant from their adjacent components and therefore form a circle in a 2 dimensional plane. Do you also think the principle components are arbitrarily located on the PCA graph of Eurogenes K15 by Orange?

You can reach the full website here: https://gen3553.pagesperso-orange.fr/ADN/K15.htm

it's not really a PCA, it's some other kind of calculation. to me personally it's better than a PCA

here you have fst distances between k15 pure admixtures
https://docs.google.com/file/d/0B9o3EYTdM8lQS3VvTUYyYXd0akk/edit

this is how it looks in PAST3
https://i.imgur.com/8vWxTqu.png

he probably made this "pca" based on these distances but i don't know how exactly.