[ncl-talk] Segmentation fault when reading variables from NC

Tue Mar 24 13:06:34 MDT 2015

I 'found' 6.2.0 and 6.2.1. I have run the attached test script with:
6.2.0, 6.2.1, 6.3.0

It runs correctly under all versions. The file is netCDF4 which,
'under-the-hood', is HDF5.

%> ncdump -k geo_em.d01.nc
netCDF-4

%> ncl_filedump geo_em.d01.nc >&! dump.geo_em

%> ncl tst_geo.ncl

===
Summary: I do not know what is wrong. I can not replicate the problem.

D

On Tue, Mar 24, 2015 at 12:13 AM, ChenFeng <fengfengabc003 at 163.com> wrote:

>
> What I do not understand is the file is generated from the standard WRF-WPS program, which I used several times with no problem.
>
> I can read the variable without the dimention name, and then reorder or subcite the data. This approach can solve the problem temporarily.
> Thanks for your help.
>
> Feng
>
>  *From:* Alex Schaefer <alexander.schaefer at mines.sdsmt.edu>
> *Date:* 2015-03-24 13:25
> *To:* ChenFeng <fengfengabc003 at 163.com>
> *CC:* Dennis Shea <shea at ucar.edu>; ncl-talk <ncl-talk at ucar.edu>
> *Subject:* Re: [ncl-talk] Segmentation fault when reading variables from
> NC
>  Feng,
>
> I got the same seg fault that you got with your file.  My guess is
> something is just not quite right with that file.  The size of that  domain
> should not be an issue.  My ncl 6.1.2 is compiled from source so the
> precompiled binary theory affecting it is not the case.  I am not sure what
> other things will spit out seg faults so Dennis will have to take lead.
>
> Sorry I couldn’t be of more help,
> -Alex
>
>
>  On Mar 23, 2015, at 9:51 PM, ChenFeng <fengfengabc003 at 163.com> wrote:
>
>  And I use the  pre-compiled binary version which is downloaded from the
> NCAR website.
>
>  *From:* ChenFeng <fengfengabc003 at 163.com>
> *Date:* 2015-03-24 11:47
> *To:* Dennis Shea <shea at ucar.edu>; alexander.schaefer
> <alexander.schaefer at mines.sdsmt.edu>
> *CC:* ncl-talk <ncl-talk at ucar.edu>
> *Subject:* Re: Re: [ncl-talk] Segmentation fault when reading variables
> from NC
>   Thanks Dennis and Alexander
>
> The reason I use the printVarSummary(lon2d) is only want to make sure whether the code above this line is correct.
> There is no problem when I use this code to read another geo_em.d01.nc
>  file. So I think there is something wrong with the NC file. But I can use ncdump to see the structure and find nothing special.
>
> Is the variable dimention too large (474*524)?  But I have also use this ncl version to read variables more large than this.
>
> I attached the file. Could you please help me check it with another ncl version and platform.
> Thanks.
>
> Feng
>
>  *From:* Dennis Shea <shea at ucar.edu>
> *Date:* 2015-03-23 22:40
> *To:* ChenFeng <fengfengabc003 at 163.com>
> *CC:* ncl-talk <ncl-talk at ucar.edu>
> *Subject:* Re: [ncl-talk] Segmentation fault when reading variables from
> NC
>  Also, forgot to mention ...
>
> Using named dimensions should only be done when reordering an array.
> Nothing wrong with doing so but is is  less efficient than ..
>
> lon2d = inp0->XLONG_M (0, :, :)
>
> On Mon, Mar 23, 2015 at 8:11 AM, Dennis Shea <shea at ucar.edu> wrote:
>
>>  I don't understand why either.  6.1.2 is no longer available here so I
>> can not test with 6.1.2
>>
>> When it does work ... what does printVarSummary(lon2d) indicate?
>>
>>  On Mon, Mar 23, 2015 at 2:47 AM, ChenFeng <fengfengabc003 at 163.com>
>> wrote:
>>
>>>   Hi ncl-users,
>>>
>>> I have encounted an error when reading a variable from an NC file.
>>> the code is listed below:
>>> ;************************************************
>>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
>>> ;************************************************
>>> begin
>>>   inp0 = addfile("/public/home/chenf/AHWRF/TEST/WPSwks.T02/geo_em.d01.nc
>>> ","r")
>>>   lon2d = inp0->XLONG_M (Time|0, south_north|:, west_east|:)
>>>   printVarSummary(lon2d)
>>> ;************************************************
>>> end
>>>
>>> the error message is:
>>>  Copyright (C) 1995-2013 - All Rights Reserved
>>>  University Corporation for Atmospheric Research
>>>  NCAR Command Language Version 6.1.2
>>>  The use of this software is governed by a License Agreement.
>>>  See http://www.ncl.ucar.edu/ for more details.
>>> Segmentation fault (core dumped)
>>>
>>>
>>> The strange thing is when i'm reading the variable without any dimention name cited, the error disappeared.
>>> The code changes into :
>>> ;************************************************
>>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
>>> ;************************************************
>>> begin
>>>   inp0 = addfile("/public/home/chenf/AHWRF/TEST/WPSwks.T02/geo_em.d01.nc
>>> ","r")
>>>   lon2d = inp0->XLONG_M
>>>   printVarSummary(lon2d)
>>> ;************************************************
>>> end
>>>
>>> I don't understand why. It is very appreciate if anybody can tell me the
>>> reason.
>>> Thanks.
>>>
>>> Feng
>>>
>>> _______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150324/91f72758/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dump.geo_em
Type: application/octet-stream
Size: 29697 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150324/91f72758/attachment.obj 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tst_geo.ncl
Type: application/octet-stream
Size: 232 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20150324/91f72758/attachment-0001.obj 