[ncl-talk] Dimension size mismatch
Rashed Mahmood
rashidcomsis at gmail.com
Sat Aug 10 03:10:27 MDT 2019
I see, since the variable " * tc1_land_MASK* " is not being reassigned or
deleted at the end of each loop iteration, I would expect that error since
your lat and lon are likely to be different in each iteration, so simplest
solution would be to use the reassignment operator ":=" for the variable
*tc1_land_MASK* and also for variable "lsmtc" .
Second option would be to delete the variables that are not necessary after
a loop iteration. e.g. just before end do line you could add:
delete([/tc1_land_MASK, lsmtc/]), you can possible add more variable to
this separated by commas. I generally prefer this second option since
deleting would release some memory next iteration.
General comment: when using loops it is important to keep in mind what
happens at successive iterations, especially when the variable dimension
sizes are expected to change in subsequent iterations.
On Sat, Aug 10, 2019 at 12:44 AM Md. Jalal Uddin <dmjalal90 at gmail.com>
wrote:
> do nr=0,nRadius-1
> do nc=0,nCenter-2
> poly_lat := clat(nc,nr,:)
> poly_lon := clon(nc,nr,:)
>
> min_lat = min(poly_lat)
> max_lat = max(poly_lat)
> min_lon = min(poly_lon)
> max_lon = max(poly_lon)
>
> ;---Extract the desired rectangle of data
>
> PRC := tcavg98({min_lat:max_lat},{min_lon:max_lon})
>
> ;---Set points that are outside of the circle of data to missing
>
> lat2d := conform(PRC, PRC&latitude, 0)
> lon2d := conform(PRC, PRC&longitude, 1)
>
> latlon_circle := gc_inout(lat2d,lon2d, poly_lat,poly_lon)
> PRC = where(latlon_circle, PRC, PRC at _FillValue)
> ;printVarSummary(PRC)
>
> if (nc.eq.0) then
>
> f_lsm = addfile("$NCARG_ROOT/lib/ncarg/data/cdf/
> landsea.nc","r")
>
> lsmtc =
> tobyte(landsea_mask(f_lsm->LSMASK,PRC&latitude,PRC&longitude))
> lsmtc at long_name = "Land-Sea Mask: TRMM 3B42"
> copy_VarCoords(PRC, lsmtc)
>
> tc_land_MASK = where(lsmtc.eq.1 .or. lsmtc.eq.2, PRC, PRC at _FillValue)
> ; Land only
> copy_VarMeta(PRC, lsmtc)
> print("tc_landgrid = "+tc_land_MASK)
>
> else
>
> f_lsm = addfile("$NCARG_ROOT/lib/ncarg/data/cdf/landsea.nc","r")
>
>
> lsmtc1 =
> tobyte(landsea_mask(f_lsm->LSMASK,PRC&latitude,PRC&longitude))
> lsmtc1 at long_name = "Land-Sea Mask: TRMM 3B42"
> copy_VarCoords(PRC, lsmtc1)
>
> * tc1_land_MASK = where(lsmtc1.eq.1 .or. lsmtc1.eq.2, PRC,
> PRC at _FillValue) ; Land only*
> copy_VarMeta(PRC, lsmtc1)
> print("tc1_landgrid = "+tc1_land_MASK)
> end if
>
> end do ; nc
> end do ; nr
>
> From the above green line error has occurred. The full code in the
> attachment.
>
> On Sat, 10 Aug 2019 at 14:27, Rashed Mahmood <rashidcomsis at gmail.com>
> wrote:
>
>> Which line the error message occurs? Generally the printVarSummary() can
>> be very useful for understanding the dimsize issues.
>>
>> On Fri, Aug 9, 2019 at 7:18 PM Md. Jalal Uddin <dmjalal90 at gmail.com>
>> wrote:
>>
>>> Thank you, sir, for your kind suggestion, but it did not work.
>>>
>>> On Fri, 9 Aug 2019 at 18:25, Rashed Mahmood <rashidcomsis at gmail.com>
>>> wrote:
>>>
>>>> Hi Jalal,
>>>> I think the error message is likely related to the lines:
>>>>
>>>> poly_lat = clat(nc,nr,:)
>>>> poly_lon = clon(nc,nr,:)
>>>>
>>>> Changed those to:
>>>> poly_lat * :=* clat(nc,nr,:)
>>>> poly_lon* :=* clon(nc,nr,:)
>>>>
>>>> hope that helps.
>>>>
>>>> Rashed
>>>>
>>>>
>>>> On Thu, Aug 8, 2019 at 6:39 PM Md. Jalal Uddin via ncl-talk <
>>>> ncl-talk at ucar.edu> wrote:
>>>>
>>>>> do nr=0,nRadius-1
>>>>> do nc=0,nCenter-1
>>>>> poly_lat = clat(nc,nr,:)
>>>>> poly_lon = clon(nc,nr,:)
>>>>>
>>>>> min_lat = min(poly_lat)
>>>>> max_lat = max(poly_lat)
>>>>> min_lon = min(poly_lon)
>>>>> max_lon = max(poly_lon)
>>>>>
>>>>> ;---Extract the desired rectangle of data
>>>>>
>>>>> PRC := jancont({min_lat:max_lat},{min_lon:max_lon})
>>>>>
>>>>> ;---Set points that are outside of the circle of data to missing
>>>>>
>>>>> lat2d := conform(PRC, PRC&latitude, 0)
>>>>> lon2d := conform(PRC, PRC&longitude, 1)
>>>>>
>>>>> latlon_circle := gc_inout(lat2d,lon2d, poly_lat,poly_lon)
>>>>> PRC = where(latlon_circle, PRC, PRC at _FillValue)
>>>>> end do
>>>>> end do
>>>>>
>>>>>
>>>>> Here, when center lat changed (nc), the dimension size of PRC also
>>>>> changed that means PRC has different size. Similar to
>>>>> http://www.ncl.ucar.edu/Applications/Scripts/polyg_29.ncl
>>>>>
>>>>> I tried to mask overland for PRC data but failed because of different
>>>>> dimension sizes.
>>>>>
>>>>> How can I make PRC as the same size? Any suggestions would be
>>>>> appreciated.
>>>>>
>>>>> Jalal
>>>>> --
>>>>> *Md. Jalal Uddin*
>>>>> MSc in Applied Meteorology
>>>>> Nanjing University of Information, Science and Technology, China
>>>>> B.Sc. in Disaster Management (Hons.)
>>>>> Patuakhali Science and Technology University, Bangladesh.
>>>>> Cell: +8613260859092, +8801792052662
>>>>> Web: www.dmjalal90.weebly.com
>>>>> Facebook: jalal.hossen.39
>>>>> LinkedIn: https://bd.linkedin.com/in/md-jalal-uddin-80a026b0
>>>>> _______________________________________________
>>>>> ncl-talk mailing list
>>>>> ncl-talk at ucar.edu
>>>>> List instructions, subscriber options, unsubscribe:
>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>
>>>>
>>>
>>> --
>>> *Md. Jalal Uddin*
>>> MSc in Applied Meteorology
>>> Nanjing University of Information, Science and Technology, China
>>> B.Sc. in Disaster Management (Hons.)
>>> Patuakhali Science and Technology University, Bangladesh.
>>> Cell: +8613260859092, +8801792052662
>>> Web: www.dmjalal90.weebly.com
>>> Facebook: jalal.hossen.39
>>> LinkedIn: https://bd.linkedin.com/in/md-jalal-uddin-80a026b0
>>>
>>
>
> --
> *Md. Jalal Uddin*
> MSc in Applied Meteorology
> Nanjing University of Information, Science and Technology, China
> B.Sc. in Disaster Management (Hons.)
> Patuakhali Science and Technology University, Bangladesh.
> Cell: +8613260859092, +8801792052662
> Web: www.dmjalal90.weebly.com
> Facebook: jalal.hossen.39
> LinkedIn: https://bd.linkedin.com/in/md-jalal-uddin-80a026b0
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20190810/6422f7bd/attachment.html>
More information about the ncl-talk
mailing list