RMallow (version 1.1)

Fit Multi-Modal Mallows' Models to Ranking Data

Description

An EM algorithm to fit Mallows' Models to full or partial rankings, with or without ties. Based on Adkins and Flinger (1998) .

Copy Link

Version

Install

install.packages('RMallow')

Monthly Downloads

138

Version

1.1

License

GPL (>= 2)

Maintainer

Erik Gregory

Last Published

February 7th, 2020

Functions in RMallow (1.1)

Rgen

Initialize sequence modes for the clustering process.

RMallow-package

Fit Multi-modal Mallows' models to ranking data.

NextTable

Calculates the table of Kendall distances in (N+1)! space, given those in N! space.

hello

Hello, World!

SimplifySequences

Change the form of ordered sequences.

UpdateLambda

Update the Lambda parameters of clusters.

SeqDistribution

Calculates distances in N! space.

Likelihood

Likelihood of the data and parameters.

AllSeqDists

Calculate all distances between a set of sequences and a fixed sequence.

three.mode

Fitted version of the toy datas data set, with three modal sequences.

datas

Sample data set.

Mallows

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

UpdateR

Update modal sequences in each cluster.

two.seq

Bi-modal Mallow's model fit to the APA data set.

two.mode

Two-mode Mallows' model fit to toy data set "datas"

UpdateP

Update Proportion in each cluster.

elect

1980 APA Presidential Candidate ranking data.

ConstructSeqs

Constructs sequences from Kendall Information matricies.

BestFit

Fit Mallows model N times and select most likely model. The EM algorithm to fit Multi-Modal Mallows' models is prone to getting stuck in local maxima, so we run it several times and selec the best one.

EStep

The Expectation step of the EM algorithm.

FormatOut

Formats the data in the "Solve" function for output.

DistanceDistribution

Calculate the Kendall distance distribution in N! space.

KendallInfo

All information used to calculate Kendall's distance.

C_lam

Calculate the normalizing coefficient for Mallow's model in a sequence space.

Lambda

Objective function to determine lambda.

AllKendall

All Kendall's distances between two sets of rankings.