simulate_tna_batches: Simulate and fit repeated TNA networks

In simulab: Simulates a Wide Variety of Data Shapes, Sizes and Distributions

Simulate and fit repeated TNA networks

Description

Generates repetitions sequence datasets with simulate_sequence_batches() and fits one transition network to each with fit_tna(), so the estimated networks can be held against the transition systems that generated them.

Usage

simulate_tna_batches(
  repetitions = 1L,
  model = c("tna", "ftna", "ctna", "atna"),
  ...,
  seed = NULL
)

Arguments

repetitions	Number of fitted networks. A single positive whole number, defaulting to 1, so a call with no arguments draws one transition system, generates sequences from it and fits one network.
model	TNA estimator, one of "tna" (the default), "ftna", "ctna" or "atna".
...	Arguments passed to simulate_sequences().
seed	Optional base seed. A single number, or NULL (the default). Dataset i uses seed + i - 1.

Value

A simulab_sim base data.frame of fitted edges, one row per dataset and from/to pair, with columns dataset, from, to and weight. Printing it shows true_transitions as the truth, keyed by the same dataset. Components: sequences, the generated sequences in long form with columns dataset, id, period and state; true_transitions, the generating probabilities with columns dataset, from, to and probability; and model_info, one row per dataset describing the fit.

Examples

if (requireNamespace("tna", quietly = TRUE)) {
  head(simulate_tna_batches(
    repetitions = 2, model = "tna", n = 20, n_states = 3, chain_length = 10, seed = 1
  ))

  # One network with no arguments, recovered against its own truth.
  network <- simulate_tna_batches(seed = 1)
  head(validate_recovery(network, network, term = "to", estimate = "weight"))
}

simulab documentation built on Oct. 15, 2026, 5:08 p.m.