cross_validate_tna: Cross-validate TNA estimators

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

Cross-validate TNA estimators

Description

Each iteration splits the sequences once into a training and a testing set, fits every estimator to both halves, and compares the two networks with compare_networks().

Usage

cross_validate_tna(
  data,
  models = c("tna", "ftna", "ctna", "atna"),
  iterations = 20L,
  training_fraction = 0.7,
  format = c("auto", "long", "wide"),
  id = "id",
  period = "period",
  state = "state",
  seed = NULL,
  ...
)

Arguments

data	Sequence data in long or wide form.
models	Character vector of TNA estimators to compare, any of "tna", "ftna", "ctna", "atna". The default runs all four.
iterations	Number of train/test splits, a whole number of at least 1, default 20.
training_fraction	Fraction of sequences assigned to the training half, strictly between 0 and 1, default 0.7. At least two training sequences are always used, and a split leaving fewer than two testing sequences is an error.
format, id, period, state	Arguments describing the shape and column names of data. They are used to reshape data before splitting; both halves are then fitted in wide form.
seed	Optional single seed, restored on exit.
...	Further arguments passed to fit_tna().

Value

A plain base data.frame with one row per iteration and estimator and the columns iteration, model, and the compare_networks() agreement metrics pearson, cosine, mae, rmse, jaccard, edges_x and edges_y. This is not a simulab_sim.

Examples

data <- simulate_sequences(n = 60, n_states = 3, chain_length = 12, seed = 1)
if (requireNamespace("tna", quietly = TRUE)) {
  cross_validate_tna(data, models = c("tna", "ftna"), iterations = 2, seed = 1)
}

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