generate_interaction_probs: generate dyadic interaction probabilities for a group with...

In EloSteepness: Bayesian Dominance Hierarchy Steepness via Elo Rating and David's Scores

generate dyadic interaction probabilities for a group with fixed individual and dyadic biases

Description

generate dyadic interaction probabilities for a group with fixed individual and dyadic biases

Usage

generate_interaction_probs(n_ind, id_bias = 0, rank_bias = 0)

Arguments

n_ind	numeric, number of individuals
id_bias	numeric, between 0 and 1. If 0 all individual are equally likely to interact. If 1, some individuals have higher propensities to interact
rank_bias	numeric, between 0 and 1. If 0 there is no relationship between rank distance and interaction propensity. If 1 there is a strong relationship: dyads closer in rank interact more often.

Value

a matrix

Examples

x <- generate_interaction_probs(n_ind = 10, id_bias = 0.2, rank_bias = 1)
rankdiff <- x[, 2] - x[, 1]
interactprob <- x[, "final"]
# closer in rank (smaller rank diff) = interaction more likely
plot(rankdiff, interactprob)

x <- generate_interaction_probs(n_ind = 10, id_bias = 0.2, rank_bias = 0)
rankdiff <- x[, 2] - x[, 1]
interactprob <- x[, "final"]
# approx. equal probs for all dyads regardless of rank diff
plot(rankdiff, interactprob)


x <- generate_interaction_probs(n_ind = 10, id_bias = 0, rank_bias = 0)
interactprob <- x[, "final"]
y <- sample(1:nrow(x), 1000, replace = TRUE, prob = interactprob)
y <- as.numeric(x[y, 1:2])
# approx. equal numbers of interactions per ID
sort(table(y))

# skewed interaction numbers
x <- generate_interaction_probs(n_ind = 10, id_bias = 1, rank_bias = 0)
interactprob <- x[, "final"]
y <- sample(1:nrow(x), 1000, replace = TRUE, prob = interactprob)
y <- as.numeric(x[y, 1:2])
sort(table(y))

EloSteepness documentation built on Sept. 21, 2023, 1:06 a.m.