in manyspectroscopicsoftwares,weconsideredthetreatedspectra
(beforedecomposition)asvariablesinPrincipalComponentAnalyses
(PCA).Asasecondapproach,the fitting parameters(bandpositions
P, fullwidthathalfmaximum W, area A, andprofile K) were
consideredasvariablesinPCA.AllthePCAscoreplotsarepresented
on thetwo first principalcomponents(PC);thefollowingPCsdonot
showamoresignificant separationbetweenthedifferentitems.More
detailed explanationsarepresentedwithintheresultspart(3.2and
3.3 paragraphs).