[ncl-talk] pdfxy/pdfxy_bin

Dennis Shea shea at ucar.edu
Fri Apr 5 13:55:37 MDT 2019


Umm, I forgot the f77 subroutine. Attached.



On Fri, Apr 5, 2019 at 6:19 AM Jonathan Buzan <jrbuzan at gmail.com> wrote:

> Hi Dennis,
>
> Thank you for this! I will fiddle around with it.
>
> -Jonathan
>
>
>
> On Apr 5, 2019, at 2:09 PM, Dennis Shea <shea at ucar.edu> wrote:
>
> Hi Jonathan,
>
> Attached is an NCL script and a  fortran-77 code from Auguat 2014. For
> some reason, these never made it into NCL. Likely no completely tested.
>
> Using the NCL function  would likely be quite slow for large arrays. Give
> it a try.
>
> %> ncl tst_pdfxyx.ncl
>
> Ypu could invoke the f77 subroutine if needed. I can't find the test for
> this. I'll look later today.
>
> %> WRAPIT xyz2pdf77.f
>
> Good luck
>
>
>
> On Thu, Apr 4, 2019 at 2:34 AM Jonathan Buzan <jrbuzan at gmail.com> wrote:
>
>> Hello NCL-Talk and Dennis Shea,
>>
>> Thanks again for the previous help, as I have had much success!
>> I am moving on to more advanced subjects. Instead of PDF(x,y), I now need
>> to calculate PDF(x,y,z).
>>
>> In the contributed.ncl file, I believe I can modify the code below by
>> adding a 3rd variable into the binning if-then-do statement. Would this be
>> along the right track?
>> I would need to create a new routine to accept the third variable in
>> the pdfxy.  Probably call it pdfxyz…
>>
>> -Jonathan
>>
>>
>> ; JRB BEGIN
>> ;print("Jon Is Here")
>> ;  pdf2 = pdfxy_bin(x,y, binxBound,binyBound, popt)
>>                           ; all NCL [original code]
>>     x1d  = ndtooned(x)  ; requires more memory
>>     y1d  = ndtooned(y)
>>
>>     do nby=0,nbinsy-1
>>        iy = ind(y1d.ge.binyBound(nby) .and. y1d.lt.binyBound(nby+1))
>>        if (.not.ismissing(iy(0))) then
>>            do nbx=0,nbinsx-1
>>               pdf2(nby,nbx) = num(x1d(iy).ge.binxBound(nbx) .and. \
>>                                   x1d(iy).lt.binxBound(nbx+1)     )
>>            end do
>>        end if
>>        delete(iy)      ; size may change
>>     end do
>>
>>
>>
>>
>>
>>
>>
>>
>> On Feb 20, 2019, at 3:30 PM, Dennis Shea <shea at ucar.edu> wrote:
>>
>> Hi Jonathan,
>>
>> No. The pdfxy/pdfxy_bin are 'cartesian oriented' (if there is such a
>> thing!).
>> Hence, x/y rathern tha lat/lon.
>>
>> D
>>
>> On Wed, Feb 20, 2019 at 6:40 AM Buzan, Jonathan <jbuzan at purdue.edu>
>> wrote:
>>
>>> Hello NCL-Talk,
>>>
>>> Do the pdfxy/pdfxy_bin functions account for lat-lon grid cell size
>>> changes? i.e. if there is no accounting for weights, the number of events
>>> at >30° are 2x to the number events at <30°, even though there are equal
>>> area.
>>>
>>> -Jonathan
>>>
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> ncl-talk at ucar.edu
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>
>> <tst_pdfxyz.ncl><PDFXYZ.ncl>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20190405/20d4f798/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: xyz2pdf77.f
Type: application/octet-stream
Size: 2791 bytes
Desc: not available
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20190405/20d4f798/attachment.obj>


More information about the ncl-talk mailing list