twitter_ABM: Agent-based model

In masonyoungblood/TwitterABM: An Agent-Based Model of Cultural Transmission on Twitter

Agent-based model

Description

An agent-based model (ABM) of cultural transmission on Twitter that incorporates content bias, frequency bias, demonstrator bias, and the level of age-dependent selection.

Usage

twitter_ABM(
  N = 1000,
  overall_activity,
  cont_bias = 0,
  dem_bias = 0,
  freq_bias = 1,
  age_dep = 1,
  obs_user_data,
  obs_init_tweets,
  sum_stats_TF = TRUE,
  diversity_TF = FALSE
)

Arguments

N	Overall population size.
overall_activity	A vector of the total of number of tweets and retweets in each timestep. The length of this vector is used to determine the number of timesteps.
cont_bias	Variation in the salience of the attractiveness of content (only positive values, where 0 is neutrality).
dem_bias	Variation in the salience of the follower count (only positive values, where 0 is neutrality).
freq_bias	Level of frequency bias (only positive values, where < 1 is novelty and > is conformity).
age_dep	Rate of decay in age-dependent selection.
obs_user_data	A data frame of the observed activity levels, probability of writing an original tweet as opposed to retweeting (mu), and the follower count for each user.
obs_init_tweets	A data frame of the observed retweet frequencies from the first timestep and the user the original tweets were from (the row of the corresponding user in obs_user_data).
sum_stats_TF	Whether you want to simplify the raw data to the following summary statistics: (1) the proportion of tweets that only appear once, (2) the proportion of the most common tweet, (3) the Hill number when q = 1 (which emphasizes more rare tweets), and (4) the Hill number when q = 2 (which emphasizes more common tweets) (TRUE/FALSE).
diversity_TF	Whether you want to return the Simpson's diversity index from each timepoint (TRUE/FALSE).

Value

Returns one object or a list of two objects, depending on the values for sum_stats_TF and diversity_TF.

masonyoungblood/TwitterABM documentation built on March 28, 2023, 3:17 p.m.