[R] STRUCCHANGE DETECTING BREAKPOINTS IN A TIME SERIES

[Achim Zeileis]

On Tue, 19 Jun 2012, denissearchundia wrote:

> HI
>
> i'm trying to detect breaks points in various flow time series, they all
> contains  seasonality and trend
> my question is :
>
> i have to remove this seasonality and trend before apply the function
> breakpoints du package strucchange??

I would recommend not removing deterministic seasonality/trend but instead 
including them in the model. The "strucchange" package contains various 
examples for this, see the various examples and accompanying papers.

For data that is not (trend-)stationary, appropriate differencing is 
typically recommended.

If the breaks in trend and the breaks in season are likely to occur at 
different times, the "bfast" package (building on "strucchange") may be 
useful in estimating more parsimonious models.

> another question, the function breakpoints is similar to de Pettit tests ?
> or how does it realy works?

I don't how the de Pettit tests so I can't say much about these. The 
methods implemented in strucchange are described in detail in various 
papers. See citation("strucchange") and help("breakpoints") for starting 
points.

hth,
Z

> THANKS!!!!
>
> DENISSE
>
> --
> View this message in context: http://r.789695.n4.nabble.com/STRUCCHANGE-DETECTING-BREAKPOINTS-IN-A-TIME-SERIES-tp4633906.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>