[−][src]Function changepoint::utils::infer_changepoints

pub fn infer_changepoints<R: Rng>(
    rs: &[Vec<f64>], 
    sample_size: usize, 
    rng: &mut R
) -> Result<Vec<f64>, CategoricalError>

Infer change-point locations from the run-length distributions

This method works by walking backwards through the run-length distributions jumping back to past distributions until the process ends at the first step.

Parameters

  • rs - Run lengths probability distributions for each step observed
  • sample_size - Number of monte-carlo draws from the rl distribution
  • rng - Random number generator

Returns

The return value is the proportion of samples which show a change-point at the given index.