Converting a timestamp to a SAS time
in [SAS]
|
Prev: online covariance-based SEM workshop using R statistical software
Next: update a column based on a column in another table
From: Arthur Tabachneck on 25 Feb 2010 12:03 Dan, I'll be the first to admit that I NEVER work with time data, but the following are the results I got when I ran the code: options datestyle=dmy; data have; informat dt1 dt2 anydtdtm17.; input dt1 dt2; format t1 t2 time8.; t1=timepart(dt1); t2=timepart(dt2); minutes=timepart(dt2)-timepart(dt1); cards; 25/01/10:07:10:25 25/01/10:07:11:25 ; proc print data=have; run; Obs dt1 dt2 t1 t2 minutes 1 1580022625 1580022685 7:10:25 7:11:25 60 Art ------- On Thu, 25 Feb 2010 08:41:45 -0800, Nordlund, Dan (DSHS/RDA) <NordlDJ(a)DSHS.WA.GOV> wrote: >Art, > >But that result was 60 seconds, not minutes, unless I am reading the time wrong. > >Dan > >Daniel J. Nordlund >Washington State Department of Social and Health Services >Planning, Performance, and Accountability >Research and Data Analysis Division >Olympia, WA 98504-5204 > > >> -----Original Message----- >> From: SAS(r) Discussion [mailto:SAS-L(a)LISTSERV.UGA.EDU] On Behalf Of >> Arthur Tabachneck >> Sent: Thursday, February 25, 2010 6:52 AM >> To: SAS-L(a)LISTSERV.UGA.EDU >> Subject: Re: Converting a timestamp to a SAS time >> >> Mark, >> >> Take a look at the example from my post earlier today. The result was 60 >> and no division was needed. >> >> options datestyle=dmy; >> data have; >> informat dt1 dt2 anydtdtm17.; >> input dt1 dt2; >> minutes=timepart(dt2)-timepart(dt1); >> cards; >> 25/01/10:07:10:25 25/01/10:07:11:25 >> ; >> >> Art >> ------ >> On Thu, 25 Feb 2010 08:48:07 -0500, Mark Miller <mdhmiller(a)GMAIL.COM> >> wrote: >> >> >Just like for a pair of SAS <date> vars or SAS <time> vars, >> >the difference between a pair of SAS <datetime> vars >> >can be computed directly by subtraction >> > i.e. <datetime1 - datetime2> >> >The difference is in seconds. >> >Divide by 60 to get minutes. >> > >> >... Mark Miller >> > >> >On Thu, Feb 25, 2010 at 7:43 AM, Arthur Tabachneck >> <art297(a)netscape.net>wrote: >> > >> >> Nick, >> >> >> >> I think you can achieve what you want with something like: >> >> >> >> options datestyle=dmy; >> >> data have; >> >> informat dt1 dt2 anydtdtm17.; >> >> input dt1 dt2; >> >> minutes=timepart(dt2)-timepart(dt1); >> >> cards; >> >> 25/01/10:07:10:25 25/01/10:07:11:25 >> >> ; >> >> >> >> HTH, >> >> Art >> >> ---------- >> >> On Feb 25, 6:14 am, nick <nickcorbin0...(a)googlemail.com> wrote: >> >> > hi, if i want to find the difference between two dates i can convert >> >> > them to SAS dates using the YYMMDD8. format and then subtract one from >> >> > the other. >> >> > >> >> > i'd like to find the time interval between two observations in >> >> > minutes. my raw data is in the format DD/MM/YYYY HH:MM:SS, i.e. >> >> > 25/01/10 07:11:25. is it possible to convert this with a format to a >> >> > SAS time? >> >> > >> >> > thanks for your help, >> >> > >> >> > nick >> >>
From: Arthur Tabachneck on 25 Feb 2010 12:07
OK .. I was definitely wrong .. those were definitely seconds and not minutes. Good thing I typically don't have to work with time data! Art -------- On Thu, 25 Feb 2010 12:03:13 -0500, Arthur Tabachneck <art297(a)NETSCAPE.NET> wrote: >Dan, > >I'll be the first to admit that I NEVER work with time data, but the >following are the results I got when I ran the code: > >options datestyle=dmy; >data have; > informat dt1 dt2 anydtdtm17.; > input dt1 dt2; > format t1 t2 time8.; > t1=timepart(dt1); > t2=timepart(dt2); > minutes=timepart(dt2)-timepart(dt1); > cards; >25/01/10:07:10:25 25/01/10:07:11:25 >; > >proc print data=have; >run; > >Obs dt1 dt2 t1 >t2 minutes > > 1 1580022625 1580022685 7:10:25 7:11:25 60 > >Art >------- >On Thu, 25 Feb 2010 08:41:45 -0800, Nordlund, Dan (DSHS/RDA) ><NordlDJ(a)DSHS.WA.GOV> wrote: > >>Art, >> >>But that result was 60 seconds, not minutes, unless I am reading the time >wrong. >> >>Dan >> >>Daniel J. Nordlund >>Washington State Department of Social and Health Services >>Planning, Performance, and Accountability >>Research and Data Analysis Division >>Olympia, WA 98504-5204 >> >> >>> -----Original Message----- >>> From: SAS(r) Discussion [mailto:SAS-L(a)LISTSERV.UGA.EDU] On Behalf Of >>> Arthur Tabachneck >>> Sent: Thursday, February 25, 2010 6:52 AM >>> To: SAS-L(a)LISTSERV.UGA.EDU >>> Subject: Re: Converting a timestamp to a SAS time >>> >>> Mark, >>> >>> Take a look at the example from my post earlier today. The result was >60 >>> and no division was needed. >>> >>> options datestyle=dmy; >>> data have; >>> informat dt1 dt2 anydtdtm17.; >>> input dt1 dt2; >>> minutes=timepart(dt2)-timepart(dt1); >>> cards; >>> 25/01/10:07:10:25 25/01/10:07:11:25 >>> ; >>> >>> Art >>> ------ >>> On Thu, 25 Feb 2010 08:48:07 -0500, Mark Miller <mdhmiller(a)GMAIL.COM> >>> wrote: >>> >>> >Just like for a pair of SAS <date> vars or SAS <time> vars, >>> >the difference between a pair of SAS <datetime> vars >>> >can be computed directly by subtraction >>> > i.e. <datetime1 - datetime2> >>> >The difference is in seconds. >>> >Divide by 60 to get minutes. >>> > >>> >... Mark Miller >>> > >>> >On Thu, Feb 25, 2010 at 7:43 AM, Arthur Tabachneck >>> <art297(a)netscape.net>wrote: >>> > >>> >> Nick, >>> >> >>> >> I think you can achieve what you want with something like: >>> >> >>> >> options datestyle=dmy; >>> >> data have; >>> >> informat dt1 dt2 anydtdtm17.; >>> >> input dt1 dt2; >>> >> minutes=timepart(dt2)-timepart(dt1); >>> >> cards; >>> >> 25/01/10:07:10:25 25/01/10:07:11:25 >>> >> ; >>> >> >>> >> HTH, >>> >> Art >>> >> ---------- >>> >> On Feb 25, 6:14 am, nick <nickcorbin0...(a)googlemail.com> wrote: >>> >> > hi, if i want to find the difference between two dates i can >convert >>> >> > them to SAS dates using the YYMMDD8. format and then subtract one >from >>> >> > the other. >>> >> > >>> >> > i'd like to find the time interval between two observations in >>> >> > minutes. my raw data is in the format DD/MM/YYYY HH:MM:SS, i.e. >>> >> > 25/01/10 07:11:25. is it possible to convert this with a format to >a >>> >> > SAS time? >>> >> > >>> >> > thanks for your help, >>> >> > >>> >> > nick >>> >> |