genotype2_ExploredModels: A data frame of competing models gathered by...

In ClustMMDD: Variable Selection in Clustering by Mixture Models for Discrete Data

Description

A data frame of competing models gathered by backward.explorer for Kmax = 10. Such data file can be used for a final model selection process.

Usage

data("genotype2_ExploredModels")

Format

A data frame with 2667 explored models on the following 16 variables.

N :

The size of the data

P :

The number of variables

K :

the number of clusters

S1 :

1st variable

S2 :

2nd variable

S3 :

3th variable

S4 :

4th variable

S5 :

5th variable

S6 :

6th variable

S7 :

7th variable

S8 :

8th variable

S9 :

9th variable

S10 :

10th variable

logLik :

The log-likelihood

dim :

The dimension = number of free parameters

entropy :

Entropy.

Details

TODO

Source

Wilson Toussile

See Also

dimJump.R and model.selection.R.

Examples

data(genotype2_ExploredModels)
head(genotype2_ExploredModels)
plot(genotype2_ExploredModels[, c("dim", "logLik")],
	 col = "blue", xlab = "Dimension", ylab = "Log-likelihood")

# Data-driven calibration of the penalty
dimJump.R(genotype2_ExploredModels, h = 5, N=1000, header=T)

ClustMMDD documentation built on May 2, 2019, 2:44 p.m.