Mallows {RMallow}

Fits a Multi-Modal Mallows' model to ranking data.

Description

Fits the Multi-Modal Mallows' model to partial or full ranking data, using Kendall's metric and an EM algorithm. This is essentially metric sequence clustering.

Usage

  Mallows(datas, G, iter = 10, hyp = NULL,
    plot.like = FALSE)

Arguments

datas

Matrix of partial or fully-ranked data.

G

Number of modes, 2 or greater.

iter

Maximum number of iterations.

hyp

Hypothesis sequence vector, to initialize one of the cluster centers at.

plot.like

Should the likelihood be printed at each iteration?

Value

See output of FormatOut

Author(s)

Erik Gregory

References

"Mixtures of distance-based models for ranking data". Thomas Brendan Murphy & Donal Martin. 1 April 2002. Computational Statistics & Data Analysis 41 (2003) 645-655.