[R] date time problem
Jeff Newmiller
jdnewmil at dcn.davis.ca.us
Mon Dec 8 09:57:13 CET 2014
You are still posting in HTML, and it is continuing to impede this
conversation. Learn how to post in plain text before posting again. Gmail
does have this option.
You are not using dput, as previously asked, either. Read the web page I
referenced to learn how to send R data unambiguously.
Your time sequence appears to include 2AM at the spring DST change... no
such point in time exists if you are working with DST, as the clock goes
from 01:59:59 to 03:00:00. I suggest that instead of "EST5EDT" you
specify timezone "Etc/GMT+5" which does not attempt to infer any DST in
the data. "EST" may also work, but there are other timezones in the world
that are referred to by that label, so it may mislead you or not work at
all in the future.
> tc <- "3/11/2007 2:00"'
> tct <- as.POSIXct( tc, format="%m/%d/%Y %H:%M", tz="Etc/GMT+5")
> tct
[1] "2007-03-11 02:00:00 GMT+5"
> as.POSIXct( tc, format="%m/%d/%Y %H:%M", tz="EST5EDT")
[1] NA
> format(tct, tz="Etc/GMT", usetz=TRUE )
[1] "2007-03-11 07:00:00 GMT"
On Sun, 7 Dec 2014, Alemu Tadesse wrote:
> Thank you very much Jeff. Below is the data I used:> Corrected_data
> SA_LST SA_GHI_mean
> 61759 3/11/2007 1:00 0.00000
> 67517 3/11/2007 2:00 0.00000
> 70017 3/11/2007 3:00 0.00000
> 70524 3/11/2007 4:00 0.00000
> 71061 3/11/2007 5:00 0.00000
> 71638 3/11/2007 6:00 0.00000
> 72113 3/11/2007 7:00 20.20873
> 72629 3/11/2007 8:00 123.14231
> 73274 3/11/2007 9:00 306.97343
>
> >test1<-as.POSIXct(Corrected_data$SA_LST,format="%m/%d/%Y %H:%M", tz="EST5EDT")
> >test1
> [1] "2007-03-11 01:00:00 EST" NA "2007-03-11 03:00:00 EDT" "2007-03-11 04:00:00 EDT"
> [5] "2007-03-11 05:00:00 EDT" "2007-03-11 06:00:00 EDT" "2007-03-11 07:00:00 EDT" "2007-03-11 08:00:00 EDT"
> [9] "2007-03-11 09:00:00 EDT"
>
> > format(test1, tz="Etc/GMT", usetz=TRUE )
> [1] "2007-03-11 06:00:00 GMT" NA "2007-03-11 07:00:00 GMT" "2007-03-11 08:00:00 GMT"
> [5] "2007-03-11 09:00:00 GMT" "2007-03-11 10:00:00 GMT" "2007-03-11 11:00:00 GMT" "2007-03-11 12:00:00 GMT"
> [9] "2007-03-11 13:00:00 GMT"
> >
>
> As you can see the "2007-03-11 02:00:00 EST" is not recognized and returned as NA. This may be due to DST, which I do not know
> how to handle in R.
>
> Best,
>
> Alemu
>
>
> On Sun, Dec 7, 2014 at 1:59 PM, Jeff Newmiller <jdnewmil at dcn.davis.ca.us> wrote:
> You have not provided a reproducible example, so anything I say could be wrong just due to misinterpretation.
> Please read [1] for suggestions on making your examples reproducible, particularly regarding the use of dput to
> provide example data. You have also posted in HTML format, which can cause additional scrambling of communications
> on this list.
>
> >From the notation you are using, I would guess that "Corrected_SA_data" is a data frame containing columns
> "date_time" and "TZ" at a minimum. The "date_time" column could be a vector of class POSIXlt or POSIXct... except
> that I cannot reproduce such an object that doesn't print some kind of timezone indicator in its output. What
> version of R are you using? Note that such an object contains a tz attribute already, so unless you have already
> made sure that the date_time column knows about that timezone, they are probably unrelated to each other.
>
> Note that any POSIXct object is internally represented as a specific instant of time. The tz attribute is only used
> to control how that time will be displayed. For example:
>
> testtime3 <- as.POSIXct( rawtime1, tz="America/New_York" )
> format( testtime3, tz="Etc/GMT", usetz=TRUE )
>
> If you want the output to omit the timezone information you can omit the usetz argument:
>
> format( testtime3, tz="Etc/GMT" )
>
> If you have not yet, you should read [2]. Note that three-letter timezone indicators (even though they are given in
> OUTPUT!) are at best unreliable for use in specifying timezones in R... read ?timezones.
>
> [1] http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example
> [2] http://www.r-project.org/doc/Rnews/Rnews_2004-1.pdf
>
> On Sun, 7 Dec 2014, Alemu Tadesse wrote:
>
> Dear R users
>
> I am puzzled by the following result from R script. I am trying to convert
> local time to UTC time. Time zone is -5, therefore I used the following
> approach.
>
> Below is the script.
> Corrected_SA_data$date_time[k-1]
>
> [1] "2007-03-11 01:00:00"
> Corrected_SA_data$TZ[k-1]
>
> [1] -5
> Corrected_SA_data$date_time[k-1]-Corrected_SA_data$TZ[k-1]*3600
>
> [1] "2007-03-11 07:00:00 MDT"
>
> I was expecting this last value to be something like "2007-03-11 06:00:00
> UTC"
>
> Please correct me if I ma wrong.
>
> On the other hand I have
> Corrected_SA_data$date_time[k]
>
> [1] "2007-03-11 02:00:00"
> Corrected_SA_data$TZ[k]
>
> [1] -5
> Corrected_SA_data$date_time[k]-Corrected_SA_data$TZ[k]*3600
>
> [1] NA
>
> I am not sure why I am getting NA.
>
> Thank you for your help.
>
> Alemu
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
>
> ---------------------------------------------------------------------------
> Jeff Newmiller The ..... ..... Go Live...
> DCN:<jdnewmil at dcn.davis.ca.us> Basics: ##.#. ##.#. Live Go...
> Live: OO#.. Dead: OO#.. Playing
> Research Engineer (Solar/Batteries O.O#. #.O#. with
> /Software/Embedded Controllers) .OO#. .OO#. rocks...1k
> ---------------------------------------------------------------------------
>
>
>
>
---------------------------------------------------------------------------
Jeff Newmiller The ..... ..... Go Live...
DCN:<jdnewmil at dcn.davis.ca.us> Basics: ##.#. ##.#. Live Go...
Live: OO#.. Dead: OO#.. Playing
Research Engineer (Solar/Batteries O.O#. #.O#. with
/Software/Embedded Controllers) .OO#. .OO#. rocks...1k
---------------------------------------------------------------------------
More information about the R-help
mailing list