Is a ts of length one a ts? (PR#245)

ripley@stats.ox.ac.uk ripley@stats.ox.ac.uk
Tue, 10 Aug 1999 18:06:15 +0200 (MET DST)


The following seems confused (0.65 snapshot)

> x <- ts(1:20)
> window(x, 1, 1)
Time Series:
Start = c(1, 1) 
End = c(1, 1) 
Frequency = 1 
[1] 1
Warning message: 
Not returning a time series object in: [.ts(x, i) 

(it of class ts, as the print method shows). Under 0.64.2 it is even more
confusing:

> x <- ts(1:20)
> window(x, 1, 1)
Warning: Not returning a time series object
Time Series:
Start = c(1, 1) 
End = c(1, 1) 
Frequency = 1 
[1] 1

sic!

> x[1]
[1] 1
Warning message: 
Not returning a time series object in: [.ts(x, 1) 

Is there any good reason why a time series of length one is not
allowed by [.ts (you can certainly create one with ts)?

How about one of length 0:

> ts(numeric(0))
Time Series:
Start = c(1, 1) 
End = c(0, 1) 
Frequency = 1 
numeric(0)

(which is surely a bug in print.ts if length 0 objects are allowed).

> x[0]
numeric(0)
Warning message: 
Not returning a time series object in: [.ts(x, 0) 

I suggest we should allow time series of length 1 but not of length 0
(as they have no start).

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._