[ncl-talk] "RasterFill" not plotting field
Mary Haley
haley at ucar.edu
Thu Jan 5 08:34:01 MST 2017
Ryan,
Well, it looks like this may have been a bug in V6.3.0, although I don't
recall anything specific to cnFillColor being fixed. This resource doesn't
get used very much.
I think this work-around may work for you: instead of setting
cnMonoFillColor and cnFillColor, set cnFillColors. You will need to define
the contour levels too, such that you have one fewer levels than you have
colors.
I've attached another version of wrf_nogsn_1.ncl that does this. Let me
know if this is not what you're looking for.
--Mary
On Wed, Jan 4, 2017 at 1:09 PM, Ryan Connelly <rconne01 at gmail.com> wrote:
> Hi Mary,
>
> opts_pv = res is just my way of initializing. There's no difference is
> switching to opts_pv = True; i.e., there isn't a hidden res causing the
> problem.
>
> I ran your script with my wrfout, but did not get your output. Instead,
> for plot 2, the masking works, but the original colors remain. I've
> attached my images. Likewise, I added the lines to turn on @RasterFill and
> @RasterSmoothing to a basic wrf_dbz plot, and that worked as expected. So
> it appears that the problem is that @cnMonoFillColor only works on contour
> plots, not on raster plots, though somehow (as your attached images show)
> this is only a problem on my end. I'm running V6.3.0 if that matters.
>
> Thanks,
> Ryan
>
>
>
> On Wed, Jan 4, 2017 at 9:46 AM, Mary Haley <haley at ucar.edu> wrote:
>
>> Ryan,
>>
>> I would need a little more information in order to debug this.
>>
>> For starters, you set "opts_pv" equal to "res". Does "res" have any
>> extra resource settings that might be affecting the contouring?
>>
>> I recommend starting with a very simple script to make sure things are
>> working. I've taken your resource settings, and created my own script that
>> plots HGT from a WRF file, and then masks some of the data values and plots
>> it again in gold, using the same settings you have. It seems to work just
>> fine for me. See attached. You should be able to run this script on your
>> own WRF output data by simply changing the name of the WRF output file in
>> the script.
>>
>> The PNG images look a little weird (shifted towards the lower right of
>> the screen) because wrf_contour is really not meant to be used as a
>> standalone function. I assume that you are eventually overlaying fill_mgpv
>> on another plot, so this shift is not an issue.
>>
>> --Mary
>>
>>
>>
>> On Tue, Jan 3, 2017 at 1:32 PM, Ryan Connelly <rconne01 at gmail.com> wrote:
>>
>>> Hi Mary,
>>>
>>> Surprisingly, it started plotting when I got into the office today. I
>>> don't know either; it must be a Christmas miracle.
>>>
>>> But now I have a new problem. I have my MGPV variable subset to just
>>> negative values using "where", and want to plot the resultant field of only
>>> negative MGPV values in gold. However, when using RasterFill, it defaults
>>> to plotting in pink, no matter what I try. I do the same process with RH,
>>> but since it is not a RasterFill, is plots in green as expected. If I add
>>> the line opts at cnFillMode = "RasterFill" to the relative
>>> humidity opts, it overrides my request and plots in clown colors, so I know
>>> that this is an error specific to RasterFill plotting. I suspect it may be
>>> a bug. Below are my opts for MGPV plotting, which I would like to appear
>>> in solid color gold:
>>>
>>> ; Plotting options for mgpv
>>> opts_pv = res
>>> opts_pv at cnLineLabelsOn = False
>>> opts_pv at cnConstFLabelPerimOn = False
>>> opts_pv at cnLabelMasking = True
>>> opts_pv at cnInfoLabelOn = False
>>> opts_pv at cnLineLabelPerimOn = False
>>> opts_pv at cnLinesOn = False
>>> opts_pv at pmLabelBarOrthogonalPosF = -0.07
>>> opts_pv at cnFillMode = "RasterFill"
>>> opts_pv at cnFillOn = True
>>> opts_pv at cnRasterSmoothingOn = True
>>> opts_pv at lbLabelBarOn = False ; Do not display label bar
>>> opts_pv at cnMonoFillColor = True
>>> opts_pv at cnFillColor = "gold"
>>> fill_mgpv = wrf_contour(a,wks,neg_mgpv,opts_pv)
>>>
>>>
>>> Thanks,
>>> Ryan
>>>
>>> On Mon, Jan 2, 2017 at 9:14 AM, Mary Haley <haley at ucar.edu> wrote:
>>>
>>>> Hi Ryan,
>>>>
>>>> You might try adding this resource, to have the raster plot generated a
>>>> slightly different way:
>>>>
>>>> res at trGridType = "TriangularMesh"
>>>>
>>>> Still, I'm curious about why the plot is not showing up with raster
>>>> fill. Is it sparse data? It would be good to get the script and data if
>>>> possible, so we can make sure it's not a bug.
>>>>
>>>> Thanks,
>>>>
>>>> --Mary
>>>>
>>>>
>>>> On Thu, Dec 29, 2016 at 1:36 PM, Ryan Connelly <rconne01 at gmail.com>
>>>> wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> I'm running a rather computationally intensive NCL script that
>>>>> attempts to calculate and plot geostrophic MPV. At times, it is
>>>>> (understandably) very *very *slow, and at other times I have
>>>>> encountered an error about workspace memory allocation being exceeded.
>>>>> With regards to the latter, I read a suggestion to change from...
>>>>>
>>>>> opts_pv at cnFillOn = True
>>>>>
>>>>> ...to...
>>>>>
>>>>> opts_pv at cnFillMode = "RasterFill"
>>>>> opts_pv at cnRasterSmoothingOn = True
>>>>>
>>>>> When doing so, the script executes much faster, but does not plot my
>>>>> desired field. (I printed out test points to confirm that the values are
>>>>> within the range I tell it to plot in ContourParameters, so I know that is
>>>>> not the problem.)
>>>>>
>>>>> Why might this be?
>>>>>
>>>>> Thanks,
>>>>> Ryan
>>>>>
>>>>> --
>>>>> Ryan Connelly
>>>>> M.S. Student in Atmospheric Sciences, Stony Brook University
>>>>> B.S. in Meteorology with Minors in Mathematics and GIS, Valparaiso
>>>>> University
>>>>> rconne01 at gmail.com
>>>>> ryan.connelly at stonybrook.edu
>>>>>
>>>>> _______________________________________________
>>>>> ncl-talk mailing list
>>>>> ncl-talk at ucar.edu
>>>>> List instructions, subscriber options, unsubscribe:
>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>
>>>>>
>>>>
>>>
>>>
>>> --
>>> Ryan Connelly
>>> M.S. Student in Atmospheric Sciences, Stony Brook University
>>> B.S. in Meteorology with Minors in Mathematics and GIS, Valparaiso
>>> University
>>> rconne01 at gmail.com
>>> ryan.connelly at stonybrook.edu
>>>
>>
>>
>
>
> --
> Ryan Connelly
> M.S. Student in Atmospheric Sciences, Stony Brook University
> B.S. in Meteorology with Minors in Mathematics and GIS, Valparaiso
> University
> rconne01 at gmail.com
> ryan.connelly at stonybrook.edu
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20170105/ab05ea50/attachment.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: wrf_nogsn_1.ncl
Type: application/octet-stream
Size: 1426 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20170105/ab05ea50/attachment.obj
More information about the ncl-talk
mailing list