factorcpt (version 0.1.2)

get.gg: Data-driven selection of the average block size

Description

Computes quantities required for data-driven selection of the average block size for stationary bootstrap

Usage

get.gg(z, M = NULL, C = 2, max.K = 5)

Arguments

z

univariate time series

M

bandwidth, by default, M = NULL to be automatically selected as in Politis and White (2004)

C

C = 2 is the default value chosen for automatic bandwidth selection in Politis and White (2004)

max.K

max.K = 5 is the default value chosen for automatic bandwidth selection in Politis and White (2004)

Value

Estimates for the two quantities required for average block size selection, see the Appendix of Barigozzi, Cho & Fryzlewicz (2016).

References

M. Barigozzi, H. Cho and P. Fryzlewicz (2016) Simultaneous multiple change-point and factor analysis for high-dimensional time series, Preprint.

D. N. Politis and H. White (2004) Automatic block-length selection for the dependent bootstrap. Econometric Reviews. 23: 53-70.