Re: Problem

Matt

I think the problem is that you are trying to read in roughly a 3.8 Gb
variable and running out of machine memory. Dennis provided a
suggestion for working around this.

What do you eventually need to do with these variables? You might be
able to subscript them in a do loop, which will be slower, but should
use less memory.

--Mary

On Fri, 12 Jun 2009, Mathew Freeman wrote:

> Mary,
>
> It gives following:
>
>
> + f=addfile("ap_0904.nc","r")
> + print(f)
> Variable: f (file variable)
>
> filename: ap_0904
> path: ap_0904.nc
> file global attributes:
> Conventions : COARDS
> calendar : standard
> comments : file created by grads using lats4d available from http://dao.gsfc.nasa.gov/software/grads/lats4d/
> model : geos/das
> center : gsfc
> history : Fri Jun 12 11:16:23 2009: ncrcat 1970.nc 1971.nc 1972.nc 1973.nc 1974.nc 1975.nc 1976.nc 1977.nc 1978.nc 1979.nc 1980.nc 1981.nc 1982.nc 1983.nc 1984.nc 1985.nc 1986.nc 1987.nc 1988.nc 1989.nc 1990.nc 1991.nc 1992.nc 1993.nc 1994.nc 1995.nc 1996.nc 1997.nc 1998.nc 1999.nc 2000.nc 2001.nc 2002.nc -o ap_0904.nc
> nco_openmp_thread_number : 1
> dimensions:
> time = 12045 // unlimited
> longitude = 360
> latitude = 220
> variables:
> double time ( time )
> units : days since 1970-1-1 0
>
> double longitude ( longitude )
> units : degrees_east
> long_name : Longitude
>
> double latitude ( latitude )
> units : degrees_north
> long_name : Latitude
>
> float precip ( time, latitude, longitude )
> comments : Unknown1 variable comment
> long_name : daily precipitation analysis interpolated into 0.25deg grids [mm/day]
> units :
> grid_name : grid-1
> grid_type : linear
> level_description : Earth surface
> time_statistic : instantaneous
> missing_value : -99.9
> _FillValue : -99.9
>
> float rstn ( time, latitude, longitude )
> comments : Unknown1 variable comment
> long_name : ratio of 0.05deg-grids with station [%]
> units :
> grid_name : grid-1
> grid_type : linear
> level_description : Earth surface
> time_statistic : instantaneous
> missing_value : -99.9
> _FillValue : -99.9
>
> + t=f->time
> + v=f->precip
> Segmentation fault
>
>
> matt
>
>
> ________________________________
> From: Mary Haley <haley_at_ucar.edu>
> To: Mathew Freeman <mattfree0180_at_yahoo.com>
> Cc: ncl-talk_at_ucar.edu
> Sent: Friday, June 12, 2009 12:50:13 PM
> Subject: Re: Problem
>
>
> On Fri, 12 Jun 2009, Mathew Freeman wrote:
>
>> Mary,
>>
>> Using "ncl -x a.ncl" without begin and end statements in script gives following lines:
>>
>> + f=addfile("ap_0904.nc","r")
>> + t=f->time
>> + v=f->precip
>> Segmentation fault
>
> How big are these variables? What does the following output:
>
> f=addfile("ap_0904.nc","r")
> print(f)
> printVarSummary(f->time)
> printVarSummary(f->precip)
>
> --Mary
>
>> The file size is 6GB. It works fine if i reduce it to 5 years to convert into months but i need a single file of all years (33 years).
>>
>> Please suggest.
>>
>> matt
>>
>>
>>
>>
>> ________________________________
>> From: Mary Haley <haley_at_ucar.edu>
>> To: mattfree0180_at_yahoo.com
>> Cc: ncl-talk_at_ucar.edu
>> Sent: Friday, June 12, 2009 12:12:08 PM
>> Subject: Re: Problem
>>
>>
>> Mathew,
>>
>> You can also locate the seg fault by commenting out the main "begin"
>> and "end" and running your script with "ncl -x your_script.ncl".
>>
>> If you are getting any error messages, it would help to include
>> these in your email message too.
>>
>> --Mary
>>
>>
>> On Fri, 12 Jun 2009, Adam Phillips wrote:
>>
>>> Hi Mathew,
>>> Can you tell us which line is causing the segmentation fault? You can
>>> use print statements to isolate the line. If the error occurs within the
>>> double do loop, let us know the i and j values when the script fails.
>>> (Again, you can use print statements for that.)
>>> Adam
>>>
>>> Mathew Freeman wrote:
>>>> Hello everyone,
>>>>
>>>> I am using the following script to convert the daily 0.25 degree data
>>>> into monthly data but getting segmentation fault error. Can anyone
>>>> please guide to this case. The script is as follows:
>>>>
>>>>
>>>> begin
>>>> f=addfile("0904.nc","r")
>>>> t=f->time
>>>> v=f->a
>>>> v2=v(latitude|:,longitude|:,time|:)
>>>> delete(v)
>>>> t=fspan(0,dimsizes(t)-1,dimsizes(t))
>>>> utc_date = ut_calendar(t, 0)
>>>>
>>>> year = floattointeger(utc_date(:,0))
>>>> month = floattointeger(utc_date(:,1))
>>>>
>>>> dims=dimsizes(v2)
>>>> v1=new((/12*(2002-1970+1),dims(0),dims(1)/),"float")
>>>> time1=fspan(0,12*(2002-1970+1)-1,12*(2002-1970+1))
>>>> time1_at_units="months since 1970-01-01"
>>>> k=0
>>>> do i=1970,2002
>>>> do j=1,12
>>>>
>>>> indx = ind(year.eq.i.and. month.eq.j)
>>>> printVarSummary(indx)
>>>> print(max(indx))
>>>> print(min(indx))
>>>> v1(k,:,:)=dim_avg(v2(:,:,indx))
>>>> delete(indx)
>>>> ; if (j.eq.1) then
>>>> print(i*100+j)
>>>> ; end if
>>>> k=k+1
>>>> end do
>>>> end do
>>>>
>>>> v1!0="time"
>>>> v1&time=time1
>>>> v1!1="lat" v1&lat=f->latitude
>>>> v1!2="lon"
>>>> v1&lon=f->longitude
>>>>
>>>> system("rm 0904_mon.nc")
>>>> fo=addfile("0904_mon.nc","c")
>>>> fo->a=v1
>>>> end
>>>>
>>>> thankyou
>>>>
>>>> matt
>>>>
>>>>
>>>>
>>>>
>>>> ------------------------------------------------------------------------
>>>>
>>>> _______________________________________________
>>>> ncl-talk mailing list
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>> --
>>> --------------------------------------------------------------
>>> Adam Phillips asphilli_at_ucar.edu
>>> National Center for Atmospheric Research tel: (303) 497-1726
>>> ESSL/CGD/CAS fax: (303) 497-1333
>>> P.O. Box 3000
>>> Boulder, CO 80307-3000 http://www.cgd.ucar.edu/cas/asphilli
>>> _______________________________________________
>>> 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 Fri Jun 12 2009 - 14:42:08 MDT