reddeer: Red Deer dataset.
In mpmm: Marginal Product Mixture Model for Relational Event Data

Description Usage Format Details References Examples

Red deer engage in aggressive acts to enforce a social hierarchy. The data consists of T = 1200 observed aggressive actions (e.g. glaring, kicking, and mounting). Each event had a clear 'winner' and 'loser', which we code as the sender and recipient, respectively.

1	data(reddeer)

A data frame with 1200 observations on the following 3 variables.

winner: a factor with levels for each deer that won
loser: a factor with levels for each deer that lost
threat1: a factor with levels CT G1 G2 MA MG MO MR SA SW XX

This is the code sheet as prepared by Michael C. Appleby to explain the data in the file REDDEER.txt. We include this discussion for reference, and we plan to include the original dataset and our processing script in later versions of the package.

—————————————————–

These are all the interactions recorded from Jan to March 1978 among stags of known identity in hard horn, i.e. with 2 normal antlers. Interactions with females, unknown males and yearlings (whose rudimentary antlers are still covered with velvet) are excluded. So are those with stags which had cast one or both antlers: this happened in older stags first, so older animals appear less and less as March progresses.

They are not a random sample of interactions from this bachelor group of stags. They were obtained during a series of watches on focal individuals, 2 of each age from 2 to 11 years old, as explained in, for example, my 1983 paper (AB 31, 913). They are thus mostly those involving the focal individual or members of his party. That same paper discussed association at some length - which affects availability to be involved in interactions.

The vast majority of these interactions were uncontested, so in a sense the fighting terminology of escalation and winning is inappropriate: hence in the explanations below I put 'winner' and 'loser' in quotes. However, we can still ask about who initiates and how far the 'winner' escalates:

Initiation: there are two ways of getting at this (as covered in the 1983 paper). First, if one opponent was moving before the interaction and the other was not, the former must have initiated. Second, for those interactions which involved a focal individual I recorded whether an approach to within 5 metres occurred; for the majority which did, the opponent which approached may be assumed to have initiated the interaction. Escalation: in the 1983 paper I classified interactions into 3 types:

1. Involving only displacement by walking towards the opponent, or very mild threats; 2. With threats such as those listed below as Group 2; 3. With threats or attacks, listed below as Group 3.

M. C. Appleby. Competition in a Red Deer Stag Social Group: Rank, Age and Relatedness of Opponents. Animal Behavior, 31:913-918, 1983.