Suggestions from both Prof Brian Ripley and Gabor Grothendiec solved
my problem.
From Prof Ripley:
>This is a problem of your OS: your example works on all of mine.
>
>Can't you change the origin in chron? If not, you should certainly be
>able to do
>
>as.POSIXct(strptime(as.character(testDATES), "%m/%d/%y"))
This still resulted in NA's on my system.
>
>Here's what I think you can do most easily:
>
>shift <- julian(1,1,1970, origin=c(month = 12, day = 30, year = 1899))
>as.POSIXct(chron(dates=unclass(testDATES) - shift))
But this worked.
From Gabor Grothendiec:
>At any rate, one idea is to create the chron dates
>relative to the default origin like this:
>
> > testDATES<-c(35947,35971,36004,36008,36053,36066)
>> testDATES.chron <- chron("12/30/1899") + testDATES
> > as.POSIXct(testDATES.chron)
>[1] "1998-05-31 20:00:00 Eastern Daylight Time"
>[2] "1998-06-24 20:00:00 Eastern Daylight Time"
>[3] "1998-07-27 20:00:00 Eastern Daylight Time"
>[4] "1998-07-31 20:00:00 Eastern Daylight Time"
>[5] "1998-09-14 20:00:00 Eastern Daylight Time"
>[6] "1998-09-27 20:00:00 Eastern Daylight Time"
>
>Does that make any difference on your system?
This worked on my machine as well. I'm not sure why these solutions
avoided the previous NA's, but I guess it had to do with my use of
the 'origin' argument within chron on my particular os...
Thanks again,
Brian
Thanks,
Brian
--
*********************************************************************
Brian Beckage
Department of Botany
University of Vermont
Marsh Life Science Building
Burlington, VT 05405
Phone: 802 656-0197
Fax : 802 656-0440
email: Brian.Beckage at uvm.edu
web : www.uvm.edu/~bbeckage
