monitor_op

docblock

monitor_op(data, lam, cl; chart_choice, m=3, d=1, add_noise=false)

Monitor the time series data with the EWMA control chart based on ordinal patterns and return a ControlChartResult with the chart statistics, the control limit and the first alarm.

The statistics are those of stat_op(data, lam; ...), and the alarm rule is the one of the chart: the chart signals when the statistic falls below cl for Shannon() and ShannonExtropy(), exceeds cl for DistanceToWhiteNoise(), and exceeds cl in absolute value for the four statistics of Bandt (2019).

data: the time series.
lam::Float64: EWMA smoothing parameter.
cl::Float64: control limit, typically obtained from cl_op.
chart_choice: one of Shannon(), ShannonExtropy(), DistanceToWhiteNoise(), UpDownBalance(), Persistence(), RotationalAsymmetry(), UpDownScaling(). For Shannon and ShannonExtropy, the statistic is in the logarithm base of the chart, which must be larger than 1. Both default to base 2 in ComplexityMeasures.jl; use Shannon(base=exp(1)) for the natural logarithm used in the papers. The control limit cl must be given in the same base.
m::Int=3: length of the ordinal patterns.
d::Int=1: delay between observations of a pattern.
add_noise::Bool=false: add uniform noise to data to break ties.

x = randn(400)
res = monitor_op(x, 0.1, 0.25; chart_choice=Persistence())
res.alarm            # position of the first signal in res.stats, or nothing