locits (version 1.4)

summary.tos: Summarize the results of a test of stationarity

Description

This function summarizes the results of a tos object that contains information about a test of stationarity. The summary function prints out information about how many individual hypothesis tests there were, how many were rejected and what the (equivalent) rejection p-value was (in the last cases both for FDR and Bonferroni). If the hypothesis of stationarity is rejected then the routine also prints out a list of the Haar wavelet coefficients (and their scales, locations and scale of the wavelet periodogram) that were significant. The function also returns a lot of this information (invisibly).

Usage

## S3 method for class 'tos':
summary(object, size = 0.05, quiet = FALSE, mctype = "FDR", ...)

Arguments

object

The output from a function that carries out a test of stationarity, e.g. hwtos2.

size

The nominal overall significance level of the test.

quiet

If this argument is TRUE then nothing is printed to the screen, although the information is still returned as an object. If FALSE then this function prints out the information about the hypothesis test.

mctype

This argument can be "FDR" for false discovery rate or "BON" for Bonferroni. This argument does not effect the basis printout. However, it does control whether FDR or Bonferroni rejects are listed, and it does control the typ

...

Other arguments

Value

• The function returns a list which contain a list of the rejected coefficients. Each list item contains the index of a particular rejected coefficient, which is a vector of at least three elements. The first element corresponds to the scale of the wavelet periodogram, the second is the level of the Haar wavelet transform, and all remaining values are the index of the significant wavelet coefficients at that Haar wavelet transform scale. The list also contains the total number of Haar wavelet coefficients rejected and the mctype argument also.

References

Nason, G.P. (2013) A test for second-order stationarity and approximate confidence intervals for localized autocovariances for locally stationary time series. J. R. Statist. Soc. B, 75, 879-904.

See Also

hwtos2, print.tos

Examples

#
# See example for hwtos2, this contains two examples where
# summary.tos (as summary(.) is used in the output.