1
$\begingroup$

I want to know that in PCA analysis or FAMD the lengths of arrows in correlation circle plot(which can be plotted by bellow code) is equal to which parameter(coefficient estimates,cos2,contribution,...) while their coordinates represent their loadings?

 fviz_pca_var(res.pca)

res.pca is the result of fitting PCA analysis on our data. enter image description here Any help would be greatly appreciated.

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ Would you kindly add the plot you are talking about to this question? $\endgroup$
    – Nikolas Rieble
    Commented Aug 28, 2017 at 13:02
  • $\begingroup$ Search pca arrows biplot and pca loading plot. This has been described many times on CV. $\endgroup$
    – ttnphns
    Commented Aug 29, 2017 at 7:39

1 Answer 1

Reset to default
-1
$\begingroup$

The length of the an arrow represents how well the parameter explains the distribution of the data. A parameter with an arrow that is clearly longer than others has a better correlation.

In your case, most arrows have a similar length, so the parameters contribute equally.

Improve this answer
$\endgroup$
2
  • 2
    $\begingroup$ A parameter with an arrow that is clearly longer than others has a better correlation That is unclear and possibly wrong. the parameter explains the distribution of the data that too. $\endgroup$
    – ttnphns
    Commented Aug 29, 2017 at 11:00
  • $\begingroup$ Perhaps supplementing this with a more quantitative explanation would help. $\endgroup$
    – Scortchi ♦
    Commented Aug 29, 2017 at 11:27

Not the answer you're looking for? Browse other questions tagged or ask your own question.