Re: [R] Strange julian and/or strptime

2008-05-23 Thread jim holtman
It is not a bug. Y2K strikes again, a couple of years late. On Fri, May 23, 2008 at 7:17 AM, Kåre Edvardsen <[EMAIL PROTECTED]> wrote: > I'm on an Ubuntu Linux PC, and get the same wrong result as you. > I could not work out what the description of '%y' really ment, so I did > not realize this w

Re: [R] Strange julian and/or strptime

2008-05-23 Thread Kåre Edvardsen
I'm on an Ubuntu Linux PC, and get the same wrong result as you. I could not work out what the description of '%y' really ment, so I did not realize this was operating system specific in this sense. Anyway, I'll find a way to work around this bug. Have a nice weekend, Kare On Fri, 2008-05-23 at

Re: [R] Strange julian and/or strptime

2008-05-23 Thread Prof Brian Ripley
On my Linux box: strptime("010169", format = "%d%m%y") [1] "1969-01-01" strptime("311268", format = "%d%m%y") [1] "2068-12-31" From the help page: '%y' Year without century (00-99). If you use this on input, which century you get is system-specific. So don't! Often values

[R] Strange julian and/or strptime

2008-05-23 Thread Kåre Edvardsen
Hi r-helpers... Why do I get this strange huge jump of 36524 days when changing "origin" from 1969-01-01 to 1968-12-31. It should still be close to zero! This really messes up my calculations of follow-up times in my analyses. > julian(strptime("010169", format = "%d%m%y"),origin = as.Date("1969-