isatloop function

Repeated Impulse Indicator Saturation

Runs isat repeatedly at pre-specified significance levels to yield multiple iterations used in

outlierscaletest.

isatloop(num=c(seq(from=20, to=1, by=-1)), t.pval.spec = FALSE,  
  print=FALSE, y, ar=NULL, iis=TRUE, sis=FALSE, ...)

Arguments

• num: numeric, target expected number of outliers under the null hypothesis, or target proportion of outliers if t.pval.spec==TRUE
• t.pval.spec: logical, if TRUE, then num specifies proportion rather than number of targeted outliers
• print: logical, if TRUE, then iterations are printed
• y: numeric vector, time-series or zoo object. Missing values in the beginning and at the end of the series is allowed, as they are removed with the na.trim command
• ar: integer vector, say, c(2,4) or 1:4. The AR-lags to include in the mean specification
• iis: logical, whether to use iis
• sis: logical, whether to use sis, default is FALSE
• ...: any argument from isat can also be used in isatloop

Details

The function repeatedly runs isat detecting outliers in a model of y at different chosen target levels of significance speciefied in num. The output of this function is used as the input for the outlierscaletest function. All additional arguments from isat can be passed to isatloop.

Returns

Returns a list of two items. The first item is the number of observations. The second item is a dataframe containing the expected and observed proportion (and number of outliers) for each specified significance level of selection.

References

Jiao, X. & Pretis, F. (2019). Testing the Presence of Outliers in Regression Models. Discussion Paper.

Pretis, F., Reade, J., & Sucarrat, G. (2018). Automated General-to-Specific (GETS) regression modeling and indicator saturation methods for the detection of outliers and structural breaks. Journal of Statistical Software, 86(3).

Author(s)

Felix Pretis, https://felixpretis.climateeconometrics.org/

See Also

isat, outlierscaletest

Examples

###Repeated isat models using the Nile dataset
  ### where p-values are chosen such that the expected number of outliers under the null
  ### corresponds to 1, 2, 3, 4 and 5.
  nile <- as.zoo(Nile)
  isat.nile.loop <- isatloop(y=nile, iis=TRUE, num=c(1,2, 3, 4, 5))