[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #ZRQ-267528]: Color bar problems

Hi Jim,

I see the issue now - thanks for the image. Just as an update, I've found where 
things appear to be going wrong in the drawing of the colorbar, and I hope to 
have a fix very soon.

Thanks for pointing this out!

Sean

> Sean:
> 
> See the attached image, which is of mean annual precipitation over
> the southwest.  It includes contours and color fil.  The color fill is
> descending.  Note how, for example, in arizona, the color fill between
> 10 and 15 inches is red, whereas on the color bar, red is between 5 and 10.
> 
> Hopefully this will show you what is going on.  The problem arises
> if you edit the color table and specify a range that is *descending*
> (e.g., 95-0) rather than *ascending* (e.g., 0-95).  I'm probably the
> only one in the world to think of doing that, but at the time, I didn't
> realize you could invert the color table.  perhaps others would try it too.
> 
> I hope this helps.
> 
> Jim
> 
> 
> On 1/27/12 2:07 PM, Unidata IDV Support wrote:
> > Greetings Jim!
> >
> > I've tried to reproduce the situation outlined in you support message, but 
> > I am just not able to do so.I may have misunderstood, but here is what I 
> > did.
> >
> > 1) I isolated two grid points in a model grid (located next to each other). 
> > Their values are 955.847 hPa and 962.87 hPa.
> >
> > 2) I set the bounds such that each grid point was contained within one 
> > increment. The upper bound was set to 962.9 hPa and the lower bound was set 
> > to 955.8 hPa, and the increment was 0.1 hPa
> >
> > 3) I changed the color scale and contour information to match the increment 
> > that contained each point (955.8 - 955.9, 962.8-962.9), and made the range 
> > such that the two values were on opposite ends of the scale. Also, only the 
> > first and last increments are set to have color (blue and red) - the middle 
> > of the range is set to white.
> >
> > 4) I switched the range from ascending to descending (from 955.8 ->  962.9 
> > to 962.9 ->  955.8). Based on your report, I expected to see one of the 
> > colored contours to turn to white on the figure, since the contours seem to 
> > shift one increment on the screen (maybe this is where I am going wrong).
> >
> >  From what I saw, everything looks ok - both colors still showed up, and 
> > nothing changed in the contouring shown on the image. I've attached two 
> > figures showing what I did.
> >
> > Perhaps I am misunderstanding the issue? Would you mind sending an example 
> > (a bundle or some screen shots?).
> >
> > Thanks!
> >
> > Sean
> >
> >>> I recently noticed an issue with the contour color fill legend *in
> >>> situations where you specify a range that is descending (i.e., from
> >>> 100-0) rather than ascending (i.e., from 0-100).  If descending, the
> >>> color fill in the figure is off by one increment compared to the
> >>> legend.  This is most easily seen if you select a color table with a
> >>> small number of increments.
> >>>
> >>> A work around is to simply invert the color table so that you can use an
> >>> ascending range, but it would be good to have this fixed so that nobody
> >>> makes an unanticipated mistake.
> >>>
> >>> Thanks,
> >>>
> >>> Jim
> >>>
> >>>
> >> Jim,
> >> It does sound like a bug.
> >> Thanks for information, we will fix this problem after the AMS.
> >>
> >> Yuan
> >
> > Ticket Details
> > ===================
> > Ticket ID: ZRQ-267528
> > Department: Support IDV
> > Priority: Normal
> > Status: Open
> 
> 
> 


Ticket Details
===================
Ticket ID: ZRQ-267528
Department: Support IDV
Priority: Normal
Status: Open