Loading ...
Sorry, an error occurred while loading the content.

108Re: date.time ??

Expand Messages
  • jbw_yes
    Feb 1, 2002
    • 0 Attachment
      Stephen,

      Cool! That's just what I needed. I understand how it works not but I
      was not sure about one part of it and that messed me up.

      Thanks again!

      - Justin

      --- In ploticus@y..., "Stephen C. Grubb" <scg@j...> wrote:
      > Justin,
      >
      > The first problem is that your data isn't truly comma delimited;
      there is
      > also whitespace between fields. In order to use delim: comma your
      data
      > needs to look like this:
      >
      > "01/29/02","16:05",68.64,205.920,13.06,55.58,
      > "01/29/02","16:10",71.19,212.858,12.92,58.27,
      > "01/30/02","16:05",68.64,205.920,13.06,55.58,
      > "01/30/02","16:10",71.19,212.858,12.92,58.27,
      > "01/31/02","16:05",68.64,205.920,13.06,55.58,
      > "01/31/02","16:10",71.19,212.858,12.92,58.27,
      >
      > It's often useful to use getdata showresults: yes to see what you
      have
      > read & parsed, when developing & debugging.
      >
      > Once your data has been cleaned up, the following script should do
      the
      > trick.. It uses a filter to combine the date & time fields. Keep
      in mind
      > that the filter result needs to use the same delimitation method as
      the
      > input (comma in this case; quotes don't matter). -Steve
      >
      > #proc getdata
      > showresults: yes
      > delim: comma
      > file: cpu.csv
      > filter:
      > ##set DT = @@1.@@2
      > ##print @DT,@@3
      >
      > // set fieldnames (post-filter)..
      > fieldnames: dt cpu
      >
      > #proc areadef
      > title: % CPU Utilization
      > titledetails: size=14 align=C
      > rectangle: 1 1 8 4
      > xscaletype: datetime mm/dd/yy.hh:mm
      > xautorange: datafield=dt
      > //xrange: 01/28/02.00:00 02/01/02.00:00
      > yrange: 0 100
      >
      > #proc xaxis
      > stubs: inc 12 hours
      > minorticinc: 30 minutes
      > stubformat: MMMdd.hhA
      >
      > #proc yaxis
      > stubs: inc 10
      > gridskip: min
      > ticincrement: 100 1000
      >
      > #proc curvefit
      > // xfield: time
      > xfield: dt
      > yfield: cpu
      > linedetails: color=blue width=.5
      > legendlabel: cpu
      >
      > #proc legend
      > location: max-1 max
      > seglen: 0.2
      >
      > ----------------
      >
      > On Thu, 31 Jan 2002, jbw_yes wrote:
      >
      > >
      > > I have a data file I am reading in and the lines look like this:
      > >
      > > "01/29/02","16:05", 68.64, 205.920, 13.06, 55.58,
      > > "01/29/02","16:10", 71.19, 212.858, 12.92, 58.27,
      > >
      > > I want to be able to do be able to graph more then one day. To do
      > > this I need to get the date and time in the date.time format. I
      have
      > > read and tried all the things I can find to do this with the
      filter:
      > > but I must be doing some thing wrong.
      > >
      > > Here is a script that works for one day but I don't know how to
      make
      > > it join the date and time with a period and use it to graph the
      data
      > > for more then one day.
      > >
      > > This works for one day:
      > >
      > > #proc page
      > > #if @DEVICE in png,gif
      > > scale: 0.8
      > > #endif
      > >
      > > #proc getdata
      > > delim: comma
      > > file: cpu.csv
      > > fieldnames: date time cpu
      > >
      > > #proc areadef
      > > title: % CPU Utilization
      > > titledetails: size=14 align=C
      > > rectangle: 1 1 8 4
      > > xscaletype: time hh:mm
      > > xrange: 00:00 23:59
      > > yrange: 0 100
      > >
      > > #proc xaxis
      > > stubs: inc 2 hours
      > > minorticinc: 30 minutes
      > > stubformat: hhA
      > >
      > > #proc yaxis
      > > stubs: inc 10
      > > gridskip: min
      > > ticincrement: 100 1000
      > >
      > > #proc curvefit
      > > xfield: time
      > > yfield: cpu
      > > linedetails: color=blue width=.5
      > > legendlabel: cpu
      > >
      > > #proc legend
      > > location: max-1 max
      > > seglen: 0.2
      > >
      > >
      > > - Justin
      > >
      > >
      > >
      > > To unsubscribe from this group, send an email to:
      > > ploticus-unsubscribe@y...
      > >
      > >
      > >
      > > Your use of Yahoo! Groups is subject to
      http://docs.yahoo.com/info/terms/
      > >
      > >
      > >
      >
      >
      > Stephen C. Grubb scg@j...
      > Scientific Software Engineer, The Jackson Laboratory
      > 600 Main Street Bar Harbor, Maine 04609 USA
    • Show all 3 messages in this topic