Hello
Can someone explain to me the diffrence between using the function "discrim" and the function "lda".
I have a dataset with 6 variables (C1, C2, ..., C6) and two groups (rn, nrn). I wante to select the variable among the 6 which influence the distinguishing between groups and estimate a linear discriminant function. We use the function lda but we don't find how to derive the : THE F test , Wilks'Lambda for each variable ,Box's M,
Eigenvalues,Canonical Correlation. Can someone tel me how we can derive the later statistics.
lda(cat ~ C1 + C2 + C3 + C4 + C5 + C6,capital0507)
Call:
lda(cat ~ C1 + C2 + C3 + C4 + C5 + C6, capital0507)
Prior probabilities of groups:
nrn rn
0.5666667 0.4333333
Group means:
C1 C2 C3 C4 C5 C6
nrn 7.399927 6.036379 7.98493 51.35301 1193.4665 116.0307
rn 13.621632 11.400917 11.05705 115.84029 883.8169 142.2850
Coefficients of linear discriminants:
LD1
C1 0.297746199
C2 -0.121145033
C3 0.930160061
C4 -0.014831134
C5 -0.002151158
C6 -0.056120761
thank you in advance
HEDHILI ZAIER LEILA
MAÏTRE ASSISTANTE EN METHODES QUANTITATIVES - INSTITUT SUPERIEUR DE GESTION-TUNIS
|