discrim
object.
discrim
object
and statistical tests of the p-variate training data used to compute
the discriminant function.
summary.discrim(object, ...)
discrim
object constructed by the
discrim
function.
summary
method.
summary.discrim
object.
Objects of this class have the following methods:
The structure of a
summary.discrim
object is a list with the following data members.
discrim
object,
object
.
method
options of the
predict
method, where the default is "plug-in".
See
predict.discrim
for details.
The error rate estimates are discussed under the
CV
data member.
classify
data member.
object
is fitted without weights using the spherical, homoscedastic, group
spherical, or heteroscedastic covariance models. See
crossvalidate.discrim
for
details on cross-validation.
Nonparametric estimates of the conditional error rates are estimated
based on miss classified counts and base on posterior probabilities.
If the prior probabilities are not proportional to the sample sizes, then stratified
posterior error rates are estimated also. As seen in the example below, it is possible
for the posterior error rate estimates to be negative.
anova.discrim
for the
training data used to compute the discriminant function.
These statistics are only computed if the spherical or homoscedastic covariance
models are used. See
anova.discrim
for details.
multicomp.discrim
. These include
Hotelling's T squared statistics for all pairs of groups in the training
data. For each significant T squared statistic
confidence intervals are computed for the differences between
means of the p-variate feature vectors of the two groups.
anova.discrim
for details).
ks.gof
for details.
Computes various statistics for the discriminant function and for the training
data used to compute the discriminant function. Most statistics computed are
implementations of methods of the
discrim
object. See the STRUCTURE section
above for the various implementations.
McLachlan, G. J. (1992).
Discriminant Analysis and Statistical Pattern Recognition,
John Wiley & Sons.
Seber, G.A.F. (1984).
Multivariate Observations,
John Wiley & Sons.
# Iris data iris.quad <- discrim(Species~Sepal.L. + Sepal.W. + Petal.L. + Petal.W., data=iris.mm, family=Classical("heter")) summary(iris.quad)