[ncl-talk] How to show P value and Trend on contour fill map?

[Atul Saini]

Thanks a lot Adam and Rashed.
Dear Adam, yours suggestion is important and works good.
Thanks for the explanation and sharing the example.

Regards,





On Sun, Dec 15, 2019 at 1:39 AM Rashed Mahmood <rashidcomsis at gmail.com>
wrote:

> Hi Adam and Atul,
> I agree that Adam's suggestion would work for this case. However, I just
> want to highlight this issue since the function trend_manken returns "0"
> for missing values. That is the reason Atul get's color over missing value
> region of variable b, since 0 is a valid number unless specified as
> missing. Similar questions have been asked previously and I think this "0"
> value return for missing input data is part of the confusion. OR, may be
> this is a feature of the function, I do not know!
>
> I wonder if the function trend_manken could return missing if a grid point
> has missing values? Below is the simplest test:
>
> print("........................... when no value is missing .......")
>  a = 1.*ispan(1,20,1)
>  b = trend_manken(a,False,0)
>  print(""+b)
>
> print("........................... when all values are missing .......")
>  c = a
>  c at _FillValue = -999.
>  c = c at _FillValue
>  d = trend_manken(c,False,0)
>  print(""+d)
>
> print("............................ when a subset of values are missing
> ..")
>  e = a
>  e at _FillValue = -999.
>  e(7)       = e at _FillValue
>  f = trend_manken(c,False,0)
>  print(""+f)
>
>
> Cheers,
> Rashed
>
>
>
>
>
>
>
> On Sat, Dec 14, 2019 at 6:46 AM Adam Phillips via ncl-talk <
> ncl-talk at ucar.edu> wrote:
>
>> Hi Atul,
>> I had a missing parentheses in my where statements as evidenced by the
>> syntax errors.
>> Change this:
>> mk1 = where(ismissing(b(0,:,:),mk1 at _FillValue,mk1)
>> to this:
>> mk1 = where(ismissing(b(0,:,:)),mk1 at _FillValue,mk1)
>> Adam
>>
>> On Dec 14, 2019, at 12:25 AM, Atul Saini <atulsainimail at gmail.com> wrote:
>>
>> 
>> Hi Dr. Adam,
>>                     I followed the way suggested by you but it didn't
>> worked and given the following syntax error. I have attached the script and
>> data file herewith this mail.
>>
>> *ncl 14>* mk1 = where(ismissing(b(0,:,:),mk1 at _FillValue,mk1)
>> fatal:syntax error: function ismissing expects 1 arguments, got 3
>> fatal:error at line 14
>> *ncl 15>* mk2 = where(ismissing(b(0,:,:),mk2 at _FillValue,mk2)
>> fatal:syntax error: function ismissing expects 1 arguments, got 3
>> fatal:error at line 15
>>
>> Regards,
>>
>>
>>
>>
>>
>> On Sat, Dec 14, 2019 at 9:41 AM Atul Saini <atulsainimail at gmail.com>
>> wrote:
>>
>>> Hi Dr. Adam,
>>>                     I followed the way suggested by you but it didn't
>>> worked. I have attached the script and data file herewith this mail.
>>>
>>> Regards,
>>>
>>>
>>>
>>>
>>>
>>> On Sat, Dec 14, 2019 at 3:37 AM Adam Phillips <asphilli at ucar.edu> wrote:
>>>
>>>> Hi Atul,
>>>> In general, NCL denotes areas that are missing graphically by coloring
>>>> them white unless modified by the user. There are no white
>>>> areas on your trend plot, and thus I am guessing that NCL believes that
>>>> the plotted grid does not have any missing data. Note that in
>>>> the trend_manken documentation that missing values are not allowed to
>>>> be input into the function. If you do have missing data
>>>> (denoted by the _FillValue attribute) within your b array, and the b
>>>> array grid points are consistently missing in time,
>>>> I would use the where function to wipe out the results from
>>>> trend_manken over those areas that were missing:
>>>> mk=trend_manken(b,opt,0)
>>>> copy_VarCoords(aveg,mk)
>>>> mk1=mk(0,:,:)
>>>> mk2=mk(1,:,:)
>>>> mk1 = where(ismissing(b(0,:,:),mk1 at _FillValue,mk1)
>>>> mk2 = where(ismissing(b(0,:,:),mk2 at _FillValue,mk2)
>>>>
>>>> The above coding assumes b at _FillValue, mk1 at _FillValue, and mk1 at _FillValue
>>>> are properly set, and that b(0,:,:) is representative of the
>>>> areas that need to be masked.
>>>>
>>>> If that advice does not fix the issue reply to the ncl-talk email list.
>>>> Adam
>>>>
>>>>
>>>>
>>>> On Thu, Dec 12, 2019 at 2:47 AM Atul Saini via ncl-talk <
>>>> ncl-talk at ucar.edu> wrote:
>>>>
>>>>> Hi NCL,
>>>>>                I have modified the script and I think I am getting the
>>>>> right plot having modified the script.
>>>>>                If, I followed the right method to plot 'probability
>>>>> level' and 'trend value', please tell me why I am getting the value for the
>>>>> region having missing values.
>>>>>
>>>>> Regards,
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Thu, Dec 12, 2019 at 2:56 PM Atul Saini <atulsainimail at gmail.com>
>>>>> wrote:
>>>>>
>>>>>> Hi NCL,
>>>>>>               I wrote the code to make map for P value and second for
>>>>>> Trend value. But, I can't understand the process of selecting P value at
>>>>>> the time of plot and Trend value.
>>>>>>
>>>>>> Please find the attached data file and script and suggest the way to
>>>>>> plot P value and Trend value.
>>>>>>
>>>>>>
>>>>>> Regards,
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>> _______________________________________________
>>>>> ncl-talk mailing list
>>>>> ncl-talk at ucar.edu
>>>>> List instructions, subscriber options, unsubscribe:
>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>
>>>>
>>>>
>>>> --
>>>> Adam Phillips
>>>> Associate Scientist,  Climate and Global Dynamics Laboratory, NCAR
>>>> www.cgd.ucar.edu/staff/asphilli/   303-497-1726
>>>>
>>>> <http://www.cgd.ucar.edu/staff/asphilli>
>>>>
>>> <MK_Test1(2).ncl>
>> <WesternAgroClimRegion_MonthlySum_1901_2017Jan.nc>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> ncl-talk at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20191215/b4bc442a/attachment.html>