Simulate behavioral data from a model

Simulate behavioral data from a model

Usage

simulate(
  structure,
  params_df,
  trial_df = NULL,
  seed = NULL,
  keep_detail = FALSE,
  keep_component = NULL,
  layout = c("auto", "rectangular", "long"),
  ...
)

# S3 method for class 'model_structure'
simulate(
  structure,
  params_df,
  trial_df = NULL,
  seed = NULL,
  keep_detail = FALSE,
  keep_component = NULL,
  layout = c("auto", "rectangular", "long"),
  ...
)

# Default S3 method
simulate(structure, ...)

Arguments

structure

Finalized model structure.

params_df

Trial-level parameter values.

trial_df

Optional data frame used to condition the simulation. When it includes an `accumulator` column, `onset` and `component` values are matched by trial and accumulator. Otherwise, values apply at the trial level.

seed

Optional random-number seed.

keep_detail

If `TRUE`, keep additional simulation detail.

keep_component

Whether to keep the chosen mixture component in the output when the model has multiple components. If `NULL`, fixed mixtures keep the component label and sampled mixtures drop it.

layout

Parameter layout. `"rectangular"` expects rows ordered by trial and accumulator. `"long"` expects rows aligned to `trial_df` or inferable from `component`. `"auto"` chooses automatically.

...

Unused; for S3 compatibility.

Value

A data frame of simulated behavioral data. For standard models this includes `trial`, `R`, and `rt`. If `n_outcomes > 1`, additional ordered response columns such as `R2`/`rt2` are included.

Examples

spec <- race_spec()
spec <- add_accumulator(spec, "A", "lognormal")
spec <- add_outcome(spec, "A_win", "A")
structure <- finalize_model(spec)
params <- c(A.m = 0, A.s = 0.1, A.q = 0, A.t0 = 0)
df <- build_param_matrix(spec, params, n_trials = 3)
simulate(structure, df, seed = 123)
#>   trial     R       rt
#> 1     1 A_win 1.083347
#> 2     2 A_win 1.168675
#> 3     3 A_win 1.131959