eloTable: Generate a series of elo ratings from interaction data

In nmmarquez/linHierarchy: Generate linear rankings from dyadic interactions within a group

Description

Given an object of the class "interData" will produce a series of elo ratings up until the latest interaction in the interData object.

Usage

eloTable(intData, initial = 1000, kScaleBounds = c(-Inf, Inf),
  kFactors = 100, minThresh = 100)

Arguments

intData	An onject of class "interData".
initial	The starting score for all participants in the group.
kScaleBounds	The player score bins to associate with K factors. See ?findK for more details
kFactors	The possible K factors to be used. See ?findK for more details.
minThresh	The minimum value of any players score. See ?eloSingleOutcome for more details

Details

Using the methods described in Neumann et al 2011. a series of Elo ratings are generated from interaction data. Initial scores may either be a single numeric value or a vector of integers correspong to the players element of intData. Further arguments are passed to the findK and singleOutcome functions and include the paramters; kScaleBounds, kFactors, and minTresh.

Value

An object of the class eloTable which contains a vector of players in the graoup, the time range of interactions, and a data frame sorted by time containing any changes in an individuals elo rating.

References

Neumann et al (2011) Assessing Dominance Hierarchies. Animal Behaviour

Examples

# generate generic data
interactions <- data.frame (a = sample (letters [1:10], 100, T),
                            b = sample (letters [1:10], 100, T),
                            o = sample (c(-1,-1,0,1,1), 100, T),
                            d = Sys.time () + runif (100, 40, 160))
# convert to interData object
id1 <- intTableConv (interactions)
# produce eloTable object
eloTable (id1)

nmmarquez/linHierarchy documentation built on May 23, 2019, 9:28 p.m.