[R] Problem with breakpoints (strucchange)

[Achim Zeileis]

On Fri, 16 Apr 2004 13:21:13 -0400 kjetil at entelnet.bo wrote:

> Hola!
> 
> I am using package strucchange, and encounters the following:
> 
> >  bp <- breakpoints(diesel90  ~ regress -1, h=NULL)
> Error in La.chol2inv(x, size) : element (14, 14) is zero, so the 
> inverse cannot be computed
> 
> The obvious problems have been checked, that is, the model matrix is 
> of full rank. What can be causing this?

For segmenting the model each sub-model you are fitting has to have a
regressor matrix of full rank. Is `regress' a factor? If so, the minimal
segment size h has to be large enough that the corresponding
coefficients can be estimated in each (potential) segment.

> I can send some data if that can be of help.

If you cannot solve the problem, you can also contact me off-list.

hth,
Z

> Kjetil Halvorsen
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html
>