Estimate movement-HMM parameters from known states

Estimate the built-in movement-HMM parameters from simulated data where latent states are observed. This helper is intended for controlled simulation studies and is not a general HMM fitting routine.

Usage

estimate_hmm_movement_oracle(
  data,
  n_states = NULL,
  initial_pseudocount = 1,
  transition_pseudocount = 1,
  min_step_variance = 1e-04,
  min_angle_sd = 0.05
)

Arguments

data

Data frame with individual_id, time, state, step_length, and turning_angle.

n_states

Number of states. If NULL, inferred from data$state.

initial_pseudocount, transition_pseudocount

Non-negative smoothing pseudo-counts.

min_step_variance

Lower bound for within-state step-length variance.

min_angle_sd

Lower bound for within-state angle standard deviation.

Value

A parameter list compatible with create_hmm_movement_parameters().