Computes bootstrap adjusted properties for control charts

Share:

Description

Computes bootstrap adjusted properties for control charts.

Usage

1
2
SPCproperty(data, nrep = 500, chart, property, params, covprob = 0.9,
  quiet = FALSE, reportdistr = FALSE, parallel = 1)

Arguments

data

The observed data.

nrep

The number of bootstrap repetitions. Default 500.

chart

The chart to be used.

property

The property to be computed. A string. Must be implemented by the chart.

params

Additional parameters for computing the property.

covprob

The coverage probability of the adjustment. Default 0.9.

quiet

Logical value indicating if progress bar should be suppressed. Default FALSE.

reportdistr

Logical value indicating if the ecdf of the bootstrap distribution should be plotted. Default FALSE.

parallel

Number of cores to use for parallel computations (using mclapply from the package parallel). Defaults to 1. If set to Inf then the number of cores is automatically detected and all but one are used. Has to be set to 1 under Windows.

Value

An object of type SPCpropertyres.

See Also

SPC2sidedconfint

Examples

1
2
3
4
5
 # calibrate CUSUM chart to an in-control ARL of 100.
 # run with a larger number of replications in real examples!
 X <-  rnorm(100) #observed data
 chart <- new("SPCCUSUM",model=SPCModelNormal(Delta=1)) # CUSUM chart with normal observations
SPCproperty(data=X,nrep=15,chart=chart,property="calARL", params=list(target=100))