hywn.test {hwwntest}

Hybrid of Box-Ljung test, Bartlett B test, Haar wavelet and General wavelet tests.

Description

Omnibus test that attempts to mitigate poor performance of single test on a particular class where it does poorly by running four different tests that work well in different directions and pooling their results.

Usage

hywn.test(x, filter.number = 10, family = "DaubExPhase")

Arguments

x

The data you wish to test (dyadic length)

filter.number

The number of vanishing moments of the wavelet used.

family

The family of the wavelets

Value

A list with components:

p.value

The overall p-value of the test

method

Text string containing the name of the method used

Author(s)

Delyan Savchev and Guy Nason

References

Nason, G.P. and Savchev, D. (2014) White noise testing using wavelets. Stat, 3, 351-362. doi:10.1002/sta4.69

See Also

bartlettB.test, genwwn.test, hwwn.test

Examples

#
# Generate test data
#
x <- rnorm(64)
#
# Run the hybrid test
#
hywn.test(x)
#
#	Hybrid Test
#
#data:  
#p-value = 0.09221