There are four small datasets to test and demonstrate EMCluster.
da1
da2
da3
myirisda1, da2, da3 are in list, and
myiris is in matrix.
da1 has 500 observations in two dimensions
da1$da$x and da1$da$y,
and they are in 10 clusters given in da1$class.
da2 has 2,500 observations in two dimensions, too. The true
parameters are given in da1$pi, da1$Mu, and da1$LTSigma.
There are 40 clusters given in da1$class for this dataset.
da3 is similar to da2, but with lower overlaps between
clusters.
myiris is selected from the original Iris dataset given by R.