R devel - Dec 2009 - cut.POSIXt problem with breaks= "quarter" (PR#14162)

Full_Name: Qing Xia
Version: 2.10.0
OS: Windows XP
Submission from: (NULL) (213.71.23.233)


I find, when I use "cut" function to cut date-time objects based on
breaks "quarter", there exists a problem. It means, if the date-time
has "JJJJ-MM-30"
or "JJJJ-MM-31", then "cut" runs not correctly, it gives
wrong levels. Detail
can be found in the following R-Output:

1. correct Output levels, if there is no date-time "JJJJ-MM-30" or
"JJJJ-MM-31"
> x<- as.Date(c("2009-03-21","2009-03-24"))
> cut(x, breaks= "quarter")
[1] 2009-01-01 2009-01-01
Levels: 2009-01-01


2. wrong Output levels, if there exists date-time "JJJJ-MM-30" or
"JJJJ-MM-31"
> x<- as.Date(c("2009-03-21","2009-03-30"))
> cut(x, breaks= "quarter")
[1] 2009-01-01 2009-01-01
Levels: 2009-01-01 2009-04-01

> x<- as.Date(c("2009-03-21","2009-03-31"))
> cut(x, breaks= "quarter")
[1] 2009-01-01 2009-01-01
Levels: 2009-01-01 2009-04-01

as.yearqtr in the zoo package could be used as a workaround until cut is fixed.
> as.yearqtr(as.Date(c("2009-03-21","2009-03-31")))
[1] "2009 Q1" "2009 Q1"
>
as.Date(as.yearqtr(as.Date(c("2009-03-21","2009-03-31"))))
[1] "2009-01-01" "2009-01-01"

On Mon, Dec 21, 2009 at 7:45 AM,  <shmunde at googlemail.com>
wrote:
> Full_Name: Qing Xia
> Version: 2.10.0
> OS: Windows XP
> Submission from: (NULL) (213.71.23.233)
>
>
> I find, when I use "cut" function to cut date-time objects based
on breaks > "quarter", there exists a problem. It means, if the
date-time has "JJJJ-MM-30"
> or "JJJJ-MM-31", then "cut" runs not correctly, it
gives wrong levels. Detail
> can be found in the following R-Output:
>
> 1. correct Output levels, if there is no date-time "JJJJ-MM-30"
or "JJJJ-MM-31"
>
>> x<- as.Date(c("2009-03-21","2009-03-24"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01
>
>
> 2. wrong Output levels, if there exists date-time "JJJJ-MM-30" or
"JJJJ-MM-31"
>
>> x<- as.Date(c("2009-03-21","2009-03-30"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01 2009-04-01
>
>
>> x<- as.Date(c("2009-03-21","2009-03-31"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01 2009-04-01
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>

What exactly is the purported bug here?  It seems to me that the cuts 
are correct and the behaiour is as documented, but that you were not 
expecting empty levels.  If that bothers you, use

cut(x, breaks= "quarter")[,drop=TRUE]


On Mon, 21 Dec 2009, shmunde at googlemail.com wrote:
> Full_Name: Qing Xia
> Version: 2.10.0
> OS: Windows XP
> Submission from: (NULL) (213.71.23.233)
>
>
> I find, when I use "cut" function to cut date-time objects based
on breaks > "quarter", there exists a problem. It means, if the
date-time has "JJJJ-MM-30"
> or "JJJJ-MM-31", then "cut" runs not correctly, it
gives wrong levels. Detail
> can be found in the following R-Output:
>
> 1. correct Output levels, if there is no date-time "JJJJ-MM-30"
or "JJJJ-MM-31"
>
>> x<- as.Date(c("2009-03-21","2009-03-24"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01
>
>
> 2. wrong Output levels, if there exists date-time "JJJJ-MM-30" or
"JJJJ-MM-31"
>
>> x<- as.Date(c("2009-03-21","2009-03-30"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01 2009-04-01
>
>
>> x<- as.Date(c("2009-03-21","2009-03-31"))
>> cut(x, breaks= "quarter")
> [1] 2009-01-01 2009-01-01
> Levels: 2009-01-01 2009-04-01
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
-- 
Brian D. Ripley,                  ripley at 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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595