normwhn.test (version 1.0)

whitenoise.test: Univariate Test for White Noise

Description

Performs an Univariate Test for White Noise. The null is white noise rather than "strict" white noise, thus permitting weak dependence in the higher moments of the variable.

Usage

whitenoise.test(x)

Arguments

x
the input is a vector of length n (observations) or a n by 1 matrix

Value

htest containing the following components:
n
no. of observations
T
length of periodogram used
MN
von Mises statistic
tMN
test statistic
test value
p-value for the test

Details

A von Mises-type statistic is computed to be valued against a N(0,4) distribution. Finite sample test statistics are thus easily generated.

References

Lobato, I., and C. Velasco (2004). "A Simple and General Test for White Noise", Econometric Society, Latin-America Meetings, Paper No. 112.