test_gop_bootstrap

Notedocblock 
test_gop_bootstrap(
  data::Vector{Float64}, n_boot::Int,
  null_dist::DiscreteUnivariateDistribution, lam::Float64;
  chart_choice, m::Int=3, d::Int=1,
  block_size::Int=1, reduce=maximum
)

Generate a bootstrap distribution of the GOP statistic. Each resampled series is monitored with the EWMA chart, and its sequence of chart statistics is summarized by reduce. Returns the vector of the n_boot resampled summaries.

  • data: The in-control time series.
  • n_boot: Number of bootstrap replications.
  • null_dist: The null (in-control) discrete distribution.
  • lam: EWMA smoothing parameter.
  • chart_choice: D_Chart() or Persistence().
  • m::Int=3: length of the ordinal patterns.
  • d::Int=1: delay between observations of a pattern.
  • block_size: Set > 1 to use block bootstrap and preserve time-series dependencies.
  • reduce=maximum: function that maps the vector of EWMA chart statistics of one resampled series to a single number.