stat_acf

Notedocblock 
stat_acf(data, h)

Compute the classical sample autocorrelation of the time series data at lag h, using StatsBase.autocor (demeaned, normalized by the full-sample variance).

  • data: the time series.
  • h::Int: lag.
Notedocblock 
stat_acf(data, lam, null_dist, acf_version)

Compute the sequence of EWMA-smoothed lag-1 autocorrelation (ACF) chart statistics for the time series data.

  • data: the time series.

  • lam::Float64: smoothing parameter of the EWMA statistic.

  • null_dist: in-control (null) distribution; its mean and variance are used to center and scale the statistic.

  • acf_version::Int: version of the ACF statistic:

    1. EWMA of the centered cross products, scaled by an EWMA of the centered squares.
    2. EWMA of the raw cross products, centered and scaled by EWMA estimates of the mean and the raw second moment.
    3. EWMA of the centered cross products, scaled by the in-control variance. Returns a vector with one chart statistic per time step.