Package index • meow

Simulation Function

The core of any individual simulation.

Load existing or simulate new data for use in a meow simulation.

data_existing(): Load data from existing files
data_simple_1pl(): A default data generation function that simulates normally distributed respondent abilities and item difficulties

Included item selection algorithms.

select_max_dist(): Item selection by network distance criterion.
select_max_dist_enhanced(): Network-based item selection with configurable edge weights.
select_max_info(): Item selection by maximum Fisher information.
select_random(): Item selection by random draw from the remaining item bank.
select_restrict_rate(): Maximum-information item selection with an exposure-rate cap.
select_sequential(): Item selection by item id, simulating a fixed test form.

Included parameter update functions. Note that some only operate on person parameters, while others simultaneously update person and item parameters.

update_maths_garden(): Elo-style updates of person and item parameters (Maths Garden).
update_prowise_learn(): Elo-style updates with paired item comparisons (Prowise Learn).
update_theta_mle(): Update person ability via maximum likelihood estimation.

Additional helper functions.

construct_adj_mat(): Construct an item-pool adjacency matrix.
meow_administered(): Logical mask of administered items.
meow_long(): Convert the matrix simulation state to a long data frame of responses.
edge_weight_inverse() edge_weight_negative_log() edge_weight_linear() edge_weight_power() edge_weight_exponential(): Alternative edge weight functions for network-based item selection