Re: argument mismatch issue

From: Coly SAR <colysar_at_nyahnyahspammersnyahnyah>
Date: Mon Oct 17 2011 - 11:41:30 MDT

Mary,

In fact, it is the kf script which requires the reodering to have time, lat
and lon. The program works fine when i use olr
 short olr(time, lat, lon)

when i use other variable as below

double pr(time, lat, lon)
 I am having the error.

Here are a few the lines of the code

CS

  ; open the file
  fileName = "clm_pr_9600.nc"
   inFile = addfile( fileName, "r" )
  obsPerDay = 1

  ; find our times
  time = inFile->time
 ; lon = inFile->lon
 ; lat =inFile->lat
  pr = inFile ->pr
  filtStart = ut_inv_calendar( 1996, 01, 01, 0, 0, 0, time@units, 0)
  filtEnd = max(time)
  ;startDate = filtStart + 192 * 24
  ;endDate = filtEnd - 192 * 24
  startDate = filtStart
  endDate = filtEnd
  ; read in the unfiltered data
 print( "Reading the data..." )
   dataShort = inFile->pr({filtStart:filtEnd},:,:)
   printVarSummary(pr)
  data = doubletofloat( dataShort )
 printVarSummary(data)
 es= data

   es(time|:,lat|y,lon|:) = (/ kf_filter( data(time|:,lat|y,lon|:), \\
         obsPerDay, 2, 10, -20, -6, 90, mis, "mrg" ) /)

******************************************************************************

2011/10/17 Mary Haley <haley@ucar.edu>

> CS,
>
> The error is telling you what the problem is. You are trying to reorder
> "data" by using named dimensions "time", "lat", "lon",
> but as you can see from the printVarSummary command, "data" doesn't have
> any named dimensions.
>
> The printVarSummary of "data" makes it look like it is already ordered time
> x lat x lon (I'm assuming this because
> the first dimension is rather large, so that's likely a time dimension, and
> the second dimension is about half the
> size of the third dimension, so this looks like lat x lon).
>
> If "data" is already time x lat x lon, then don't use dimension
> reordering:
>
> es(time|:,lat|y,lon|:) = (/ kf_filter( data,obsPerDay, 2, 10, -20, -6,
> 90, mis, "mrg" ) /)
>
> I don't know why you have the reordering on the left side as well. Are you
> sure you need to do this?
>
> --Mary
>
> On Oct 16, 2011, at 10:18 AM, Coly SAR wrote:
>
> I had a progress but having an error in variable data with no specified
> time , lat ,lon
> ............................
> Variable: data
> Type: float
> Total Size: 75221244 bytes
> 18805311 values
> Number of Dimensions: 3
> Dimensions and sizes: [1827] x [73] x [141]
> Coordinates:
> Number Of Attributes: 1
> _FillValue : -9.99e+08
> (0) Filtering...dim oct 16 16:13:54 GMT 2011
> fatal:(lat) is not a named dimension in variable (data).
> fatal:Execute: Error occurred at or near line 83
>
> CS
>
> 2011/10/16 Dennis Shea <shea@ucar.edu>
>
>> This type of error can be diagnosed by doing a printVarSummary(...)
>> of kf_filter and es. Or, print(typeof(kf_filter)) and print(typeof(es))
>>
>> Perhaps, kf_filter is type double and es is type float? If so,
>>
>> es(time|:,lat|y,lon|:) = tofloat( \
>>
>> (/ kf_filter( data(time|:,lat|y,lon|:), \
>> obsPerDay, 2, 10, -20, -6, 90, mis, "mrg" ) /) )
>>
>>
>> On 10/16/11 9:08 AM, Coly SAR wrote:
>>
>>> Dear all,
>>>
>>> I am having a mismatch argument error when running a script with my data
>>> as shown below:
>>>
>>> Number of Dimensions: 3
>>> Dimensions and sizes: [time | 1827] x [lat | 73] x [lon | 141]
>>> Coordinates:
>>> time: [ 0..2629440]
>>> lat: [-0.5..35.5]
>>> lon: [-22.5..47.5]
>>> Number Of Attributes: 1
>>> _FillValue : -999000000
>>> (0) 0 72 dim oct 16 14:55:40 GMT 2011
>>> fatal:Argument type mismatch on argument (0) of (kf_filter) can not
>>> coerce
>>> fatal:Execute: Error occurred at or near line 97
>>>
>>> Hoping to have a hint to solve this urgent problem.
>>>
>>> The line of the code is below:
>>>
>>> es(time|:,lat|y,lon|:) = (/ kf_filter( data(time|:,lat|y,lon|:), \\
>>> obsPerDay, 2, 10, -20, -6, 90, mis, "mrg" ) /)
>>>
>>> Thanks,
>>>
>>> CS
>>>
>>>
>>>
>>> ______________________________**_________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/**mailman/listinfo/ncl-talk<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
>
>
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Oct 17 11:41:38 2011

This archive was generated by hypermail 2.1.8 : Tue Oct 18 2011 - 09:42:15 MDT