fstat {spatclus}R Documentation

Test statistic

Description

Computes the F-statistics for each model with $k$ breaks for $k in 1,...,m$, and the WD max F-statistic that allows to select the number of breaks and determines if the cluster(s) is significant.

Usage

 fstat(reslst,bc,m,T,eps)

Arguments

reslst A matrix giving, for each point ordered by its rank in the trajectory, its distance to the nearest neighbourg, the expentancy of this distance, and its weighted distance.
bc A list. The element $k$ of the list is a vector containing the estimated breaks for the $k$-breaks model.
m The maximum number of breaks.
T The size of the weighted serie.
eps Minimum size of cluster (ratio of the total number of cases).

Details

The theory used in this function is based on works of J. Bai and P. Perron.

Value

A list of objects :

F A F-statistics matrix giving in the first row the F-statistic value for each of the $m$ models, and the corresponding weighted F-statistics in the second row.
wdm The WD max F-statistic value.
wdms The WD max F-statistic threshold at $5%$.
signif A boolean. If True, the cluster(s) located is significant.
kmax The number of breaks of the model maximizing the WD max F-statistic.

Author(s)

Christophe Dematte{"i} demattei@iurc.montp.inserm.fr

References

Bai J. and Perron P. (1998), Estimating and testing linear models with multiple structural changes. Econometrica, 66, 47–78.

Bai J. and Perron P. (2003), Critical values for multiple structural change tests. Econometrics Journal, 6, 72–78.

See Also

supf critval critvalwdm


[Package spatclus version 1.0-3 Index]