[R] Subsetting Time Series

[Martin Maechler]

>>>>> "Richard" == Richard A Bilonick <rab at nauticom.net>
>>>>>     on Thu, 17 Apr 2003 12:23:56 -0400 writes:

   Richard> Martin Maechler wrote:

  >> It seems you want to  *replace*  rather than  *subset* 
  >> (otherwise, try to be more specific), and 
  >> there's no problem with that, e.g., with first two lines from
  >> example(ts) :
  >> 
  >> 
  >> 
  >> > gnp <- ts(cumsum(1 + round(rnorm(100), 2)), start = c(1954,7), frequency= 12)
  >> > plot(gnp)
  >> > gnp[20] <- 80
  >> > str(gnp)
  >> Time-Series [1:100] from 1954 to 1963: 2.47 3.26 4.50 5.18 6.31 ...
  >> > lines(gnp, col=2)
  >> 
  >> ------------
  >> 
  >> If you really want to subset, you can use  window(.) on your
  >> time series, but only for those kinds of subsetting.
  >> General subsetting doesn't give regularly spaced series.
  >> 
  >> {"thinning" (e.g. taking every 2nd value) would be one kind of
  >> subsetting that could be made to work ... 
  --> proposals please to R-devel at lists.R-project.org :-) 
  >> }

    Richard> There does seem to be a problem in 1.6.2 - or am I
    Richard> missing something?  First, x is a ts object. Once I
    Richard> use subsetting notation to replace one value, x
    Richard> does not seem to print correctly, and is.ts reports
    Richard> FALSE:

    >> x <- ts(1:12,start=c(2003,1),frequency=12)
    >> x
    Richard> Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
    Richard> 2003   1   2   3   4   5   6   7   8   9  10  11  12
    >> x[1]
    Richard> [1] 1
    >> x[1] <- 9
    >> x
    Richard> [1]  9  2  3  4  5  6  7  8  9 10 11 12
    Richard> attr(,"tsp")
    Richard> [1] 2003.000 2003.917   12.000
    Richard> attr(,"class")
    Richard> [1] "ts"
    >> x
    Richard> [1]  9  2  3  4  5  6  7  8  9 10 11 12
    Richard> attr(,"tsp")
    Richard> [1] 2003.000 2003.917   12.000
    Richard> attr(,"class")
    Richard> [1] "ts"
    >> as.ts(x)
    Richard> Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
    Richard> 2003   9   2   3   4   5   6   7   8   9  10  11  12
    >> is.ts(x)
    Richard> [1] FALSE

    Richard> Is it supposed to work this way?

No.  Yes, there's a bug in there, 
but only when the time series is constructed from an *integer*
vector -- which is very rare in practice but happens in these
toy examples (and we had one report of practical data !)
Anyway,
     ---> your personal first reason to upgrade to version 1.7.0 !

If you use   as.numeric(1:12)    instead of   1:12
in your examples, all is well, even for R versions prior to 1.7

Martin Maechler <maechler at stat.math.ethz.ch>	http://stat.ethz.ch/~maechler/
Seminar fuer Statistik, ETH-Zentrum  LEO C16	Leonhardstr. 27
ETH (Federal Inst. Technology)	8092 Zurich	SWITZERLAND
phone: x-41-1-632-3408		fax: ...-1228			<><