Fwd: Re: clmMonTLL produces nan

Neglected to cc ncl-talk

-------- Original Message --------
Subject: Re: clmMonTLL produces nan
Date: Thu, 02 Aug 2012 09:50:37 -0600
From: Dennis Shea <shea@ucar.edu>
To: Christopher Danek <m0zi@gmx.de>

Not sure what has happened. For sure, this is an issue

> missing_value : 1e+20
> _FillValue : -999

They should be the same.
      psli@_FillValue = psli@missing_value

missing_value is not recognized by NCL ... only _FIllValue

===============================

psli = f->PSLI

(1) print("psli: min="+min(psli)+" max="+max(psli))

(2) Does the original input have NaNs?

     knt_NaN = num(isnan_ieee(psli)) ; count the # of Nan
     print("knt_NaN="+knt_NaN)

     if (knt_NaN.gt.0) then
       replace_ieeenan (psli, psli@missing_value, 0)
       psli@_FillValue = psli@missing_value
     end if

On 08/02/2012 09:25 AM, Christopher Danek wrote:
> Hi all
>
> I have an input dataset of sea level pressure data with named coords
> (time, lat, lon) as desired for the function clmMonTLL:
>
> Variable: psli
> Type: float
> Total Size: 28911504 bytes
> 7227876 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 2676] x [lat | 37] x [lon | 73]
> Coordinates:
> time: [15.5..81433.5]
> lat: [-90..90]
> lon: [-180..180]
> Number Of Attributes: 9
> lonFlip : longitude coordinate variable has been reordered via lonFlip
> history : At 00:50:53 on 04/22/2005: CMOR altered the data in
> the following ways: replaced missing value flag (1.00000E+28) with
> standard missing value (1.00000E+20);
> original_name : mslp
> missing_value : 1e+20
> _FillValue : -999
> cell_methods : time: mean
> units : Pa
> long_name : Sea Level Pressure
> standard_name : air_pressure_at_sea_level
>
> I can plot the data and everything is fine.
> Now all I do is this:
>
> pslclm = clmMonTLL(psli)
>
> Produces the output:
>
> Variable: pslclm
> Type: float
> Total Size: 129648 bytes
> 32412 values
> Number of Dimensions: 3
> Dimensions and sizes: [month | 12] x [lat | 37] x [lon | 73]
> Coordinates:
> month: [0..11]
> lat: [-90..90]
> lon: [-180..180]
> Number Of Attributes: 11
> lonFlip : longitude coordinate variable has been reordered via lonFlip
> history : At 00:50:53 on 04/22/2005: CMOR altered the data in
> the following ways: replaced missing value flag (1.00000E+28) with
> standard missing value (1.00000E+20);
> original_name : mslp
> missing_value : 1e+20
> _FillValue : -999
> cell_methods : time: mean
> units : Pa
> long_name : Sea Level Pressure
> standard_name : air_pressure_at_sea_level
> time_op_ncl : Climatology: 223 years
> info : function clmMonLLT: contributed.ncl
>
> So far so good. But if I look at pslclm, every value became a "nan" or
> "-nan" like this:
> ...
> (11,31,69) -nan
> (11,31,70) nan
> (11,31,71) nan
> (11,31,72) -nan
> (11,32,0) -nan
> (11,32,1) nan
> (11,32,2) nan
> ...
>
> The former FillValues are still FillValues.
>
> I didn't find a solution for this problem although I'm sure it's
> something stupid.
> Thanks for help!
> Best regards
> Chris
>
>
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Thu Aug 2 09:53:27 2012