Hi there,
My name is Renan X. Cortes, student of Statistics, from south of Brazil, and I'd
like to ask you a few questions about decomposition of time series.
In R, when I fit the decomposition using the "stl" function, an
object is returned when ask the summary of the fit, called STL.seasonal (%),
STL.trend (%) and STL.remainder (%).
Once the decomposition is additive, I thought that this would be some kind of
decompositon of the variability of the time serie in terms of seasonal, trend
and residual unexplained. Just like a factorial analysis.
But, the sum os the %'s isn't one. In fact, in some cases the value of the
STL.seasonal or STL.trend exceeds 100%.
When I read the paper of the help of the function "STL: A
Seasonal-Trend Decomposition Procedure Based on Loess", I've seen
that the % of the components due the decomposition was constructed under
the eigenvalues of the operators matrices T and S. But It's not clear for me,
in the stl function in R
what exactly does these %'s means.
What does these values means? Is there a practical interpretation for them? If
so, which is?
I attached a file showing the values.
This doubt is killing my nights of sleep.
Best regards,
Renan Xavier Cortes
Departament os Statistics
Universidade Federal do Rio Grande do Sul, Brasil
_________________________________________________________________
Facebook.
cial-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_2:092
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
