Using R and RStudio for Data Management, Statistical Analysis, and Graphics, 2nd Edition

i

i

“K23166” — 2015/1/28 — 9:35 — page 121 — #147

i

i

i

i

i

i

7.10. EXAMPLES 121

> library(MASS)

> ngroups = length(unique(ds$homeless))

> ldamodel = lda(homeless ~ age + cesd + mcs + pcs,

prior=rep(1/ngroups, ngroups), data=ds)

> print(ldamodel)

Call:

lda(homeless ~ age + cesd + mcs + pcs, data = ds, prior = rep(1/ngroups,

ngroups))

Prior probabilities of groups:

0 1

0.5 0.5

Group means:

age cesd mcs pcs

0 35.0 31.8 32.5 49.0

1 36.4 34.0 30.7 46.9

Coefficients of linear discriminants:

LD1

age 0.0702

cesd 0.0269

mcs -0.0195

pcs -0.0426

The results indicate that homeless subjects tend to be older, have higher CESD scores,

and lower MCS and PCS scores. Figure 7.4 displays