How significant a PCA separation is?

[rozitaa]

Hi,

I have done a PCA analysis and now I want to get a p-value for my group separation to know if this classification is significant and/or between which categorical variables it is significant.
It would be also great if I can find out based on which dimension I can get the maximum separation for each categorical variable.

Thank!

[bioinfobeginner]

Hi,

I have done a PCA analysis and now I want to get a p-value for my group separation to know if this classification is significant and/or between which categorical variables it is significant.
It would be also great if I can find out based on which dimension I can get the maximum separation for each categorical variable.

Thank!

Hi rozitaa,
Did u find answer to your question. I am also looking for something like this. Please let me know if you know something.

Thanks!

[Dario1984]

Principal Component Analysis is an exploratory data analysis technique. It does not test a null hypothesis and generate a p-value.

[rozitaa]

Hi everyone,

I have found a package in R called, FactoR which seems it can give you a p-value for component separation. I haven't tested it myself yet but please have a look and I am looking forward to hearing your opinion about it.

http://factominer.free.fr/classical-methods/principal-components-analysis.html

[JackieBadger]

FactoR mine is essentially performing independent comparisons among the data here like one would with a correlation coefficient or regression. Strictly speaking it is not giving a P value for the PCA. These are ad hoc comparisons to help interpret the PCA.