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

Re: [ploticus] major bug with datetime x values?

Expand Messages
  • Stephen C. Grubb
    Jason, You can get around this problem by using: xautorange: datafield=1 nearest=hour Meanwhile, you are right, something is not quite correct.. proc lineplot
    Message 1 of 4 , Aug 22, 2002
    • 0 Attachment
      Jason,

      You can get around this problem by using:

      xautorange: datafield=1 nearest=hour

      Meanwhile, you are right, something is not quite correct.. proc lineplot
      apparently thinks that the extreme points are out of the plotting area. I
      don't know that I'd call it a "major" bug because in most cases you would
      want to use a nearest= attribute on autorange when dealing with datetime
      data. I'll check into it.

      -Steve


      On Mon, 12 Aug 2002, Jason wrote:

      > howdy
      >
      > I think I've dug up a significant bug. Ploticus seems to not plot some points
      > depending on the seconds field of a datetime. To illustrate, here's a
      > simple prefab:
      >
      > //////////////// start prefab
      > #proc getdata
      > standardinput: yes
      > delim: space
      > #proc endproc
      >
      > #proc areadef
      > rectangle: 1 2 5.5 4
      > xscaletype: datetime yyyy-mm-dd.hh:mm:ss
      > xautorange: datafield=1
      > yautorange: datafield=2
      > #proc endproc
      >
      > #proc lineplot
      > xfield: 1
      > yfield: 2
      > #proc endproc
      > //////////////// end prefab
      >
      > now use it to plot this data and all's well:
      > 2002-08-08.12:00:00 10
      > 2002-08-08.14:00:00 11
      > 2002-08-08.16:00:00 10
      > 2002-08-08.18:00:00 11
      > 2002-08-08.20:00:00 10
      >
      > but use it to plot this data and it drops the last point:
      > 2002-08-08.12:00:00 10
      > 2002-08-08.14:00:00 11
      > 2002-08-08.16:00:00 10
      > 2002-08-08.18:00:00 11
      > 2002-08-08.20:00:01 10
      >
      > plot this data and it drops first and last:
      > 2002-08-08.12:00:59 10
      > 2002-08-08.14:00:00 11
      > 2002-08-08.16:00:00 10
      > 2002-08-08.18:00:00 11
      > 2002-08-08.20:00:01 10
      >
      > What's the trick to getting around this problem?
      >
      > --
      > Jason Burnett
      > jason@...
      > ~ No witty signature available at this time ~
      >
      >
      > To unsubscribe from this group, send an email to:
      > ploticus-unsubscribe@yahoogroups.com
      >
      >
      >
      > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
      >
      >
      >


      Stephen C. Grubb scg@...
      Scientific Software Engineer, The Jackson Laboratory
      600 Main Street Bar Harbor, Maine 04609 USA
    • Stephen C. Grubb
      Jason, I have looked into this further and found that the problem is caused by rounding error introduced during datetime conversion, and proc lineplot
      Message 2 of 4 , Aug 22, 2002
      • 0 Attachment
        Jason,

        I have looked into this further and found that the problem is caused by
        rounding error introduced during datetime conversion, and proc lineplot
        determining that the extreme points were just out of range (proc lineplot
        doesn't render points outside the X range). I have added a new lineplot
        attribute that relaxes this for the occasional times when it is useful; it
        will be present in next release.

        -Steve


        On Thu, 22 Aug 2002, Stephen C. Grubb wrote:

        > Jason,
        >
        > You can get around this problem by using:
        >
        > xautorange: datafield=1 nearest=hour
        >
        > Meanwhile, you are right, something is not quite correct.. proc lineplot
        > apparently thinks that the extreme points are out of the plotting area. I
        > don't know that I'd call it a "major" bug because in most cases you would
        > want to use a nearest= attribute on autorange when dealing with datetime
        > data. I'll check into it.
        >
        > -Steve
        >
        >
        > On Mon, 12 Aug 2002, Jason wrote:
        >
        > > howdy
        > >
        > > I think I've dug up a significant bug. Ploticus seems to not plot some points
        > > depending on the seconds field of a datetime. To illustrate, here's a
        > > simple prefab:
        > >
        > > //////////////// start prefab
        > > #proc getdata
        > > standardinput: yes
        > > delim: space
        > > #proc endproc
        > >
        > > #proc areadef
        > > rectangle: 1 2 5.5 4
        > > xscaletype: datetime yyyy-mm-dd.hh:mm:ss
        > > xautorange: datafield=1
        > > yautorange: datafield=2
        > > #proc endproc
        > >
        > > #proc lineplot
        > > xfield: 1
        > > yfield: 2
        > > #proc endproc
        > > //////////////// end prefab
        > >
        > > now use it to plot this data and all's well:
        > > 2002-08-08.12:00:00 10
        > > 2002-08-08.14:00:00 11
        > > 2002-08-08.16:00:00 10
        > > 2002-08-08.18:00:00 11
        > > 2002-08-08.20:00:00 10
        > >
        > > but use it to plot this data and it drops the last point:
        > > 2002-08-08.12:00:00 10
        > > 2002-08-08.14:00:00 11
        > > 2002-08-08.16:00:00 10
        > > 2002-08-08.18:00:00 11
        > > 2002-08-08.20:00:01 10
        > >
        > > plot this data and it drops first and last:
        > > 2002-08-08.12:00:59 10
        > > 2002-08-08.14:00:00 11
        > > 2002-08-08.16:00:00 10
        > > 2002-08-08.18:00:00 11
        > > 2002-08-08.20:00:01 10
        > >
        > > What's the trick to getting around this problem?
        > >
        > > --
        > > Jason Burnett
        > > jason@...
        > > ~ No witty signature available at this time ~
        > >
        > >
        > > To unsubscribe from this group, send an email to:
        > > ploticus-unsubscribe@yahoogroups.com
        > >
        > >
        > >
        > > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
        > >
        > >
        > >
        >
        >
        > Stephen C. Grubb scg@...
        > Scientific Software Engineer, The Jackson Laboratory
        > 600 Main Street Bar Harbor, Maine 04609 USA
        >
        >


        Stephen C. Grubb scg@...
        Scientific Software Engineer, The Jackson Laboratory
        600 Main Street Bar Harbor, Maine 04609 USA
      • Jason
        ... Thanks, Steve. Great software, great support, and great price :) -- Jason Burnett jason@jnj.org ~ No witty signature available at this time ~
        Message 3 of 4 , Aug 22, 2002
        • 0 Attachment
          On Thu, Aug 22, 2002 at 10:30:31AM -0400, Stephen C. Grubb wrote:
          > Jason,
          >
          > I have looked into this further and found that the problem is caused by
          > rounding error introduced during datetime conversion, and proc lineplot
          > determining that the extreme points were just out of range (proc lineplot
          > doesn't render points outside the X range). I have added a new lineplot
          > attribute that relaxes this for the occasional times when it is useful; it
          > will be present in next release.
          >

          Thanks, Steve. Great software, great support, and great price :)

          --
          Jason Burnett
          jason@...
          ~ No witty signature available at this time ~
        Your message has been successfully submitted and would be delivered to recipients shortly.