[Paraview] [EXTERNAL] Paraview can't show grads origin data

[Moreland, Kenneth]

That crash sounds like it could be a bug.  I vaguely recall recent similar
reports about crashes and animations, but I have not been following them.
Perhaps someone else on the list can report.


Is there any chance you can update to the latest version of ParaView
(3.14)?  I wouldn't want anyone chasing a bug that has already been fixed.

-Ken

On 3/15/12 2:52 AM, "wang" <wang at knmi.nl> wrote:

>Dear Kenneth,
>
>Thanks a lot for you effort to reply my mail.
>
>The problem is solve now. It was simpler than it looks. It only because
>the data produced in this way is 4-D (x,y,z,t). When I convert it into
>3-D(x,y,t)
>everything is fine.
>
>But I have some other problems: the Paraview *crashes* when I'm making
>movie
>or just make many png files series. I was forced to cut the time steps
>into many
>small pieces in order to keep Paraview running. It makes no difference
>whether
>use 1fps or 2**n fps.
>
>A few months ago I could make avi formatted movie without any problem. And
>in the time nothing has  changed with my computer. By the way I'm using
>version 3.10.
>
>Cheers!
>Xueli
>
>
>Moreland, Kenneth wrote:
>> OK.  I see now.  Clearly I'm not familiar with the GrADS software or its
>> data formats.  If someone else on the ParaView mailing list is familiar
>> with GrADS it would be great if they speak up, but according to the mail
>> archives GrADS has not yet been discussed.
>>
>>
>> I suspect the problem is that ParaView simply does not know how to read
>> the type of netCDF files written by GrADS.  NetCDF is not a completely
>> specified file format.  Rather, it is a collection of arrays that are
>> "self describing."  That is, the arrays and dimensions are named and can
>> have properties, but to read them properly you really have to understand
>> what these metadata mean.
>>
>> There are actually several readers in ParaView that use netCDF as the
>> underlying technology, but only one that makes an attempt to generally
>> read unknown "conventions."  It will assume that arrays are regular
>>grids
>> with up to three physical dimensions (x/y/z or lat/lon/z) and
>>optionally a
>> time dimension.  Based on about 5 minutes looking at the GrADS web
>>pages,
>> I'm guessing that the netCDF file GrADS is outputting contains 5
>> dimensional arrays.  In this case, the ParaView reader won't know what
>>to
>> do with 5 dimensions and will therefore issue an error if you try to
>>read
>> one.  (You never actually said what went wrong, but perhaps that is it.)
>>
>> I'm sorry that probably none of this is helping.  Ultimately, it sounds
>> like a new reader would have to be created for ParaView.
>>
>> -Ken
>>
>> On 3/13/12 2:28 AM, "wang"<wang at knmi.nl>  wrote:
>>
>>    
>>> Thanks for your reply.
>>>
>>> Grads is a software package that widely used by meteorological
>>>societies
>>>   worldwide. http://www.iges.org/grads/.
>>> The nc file I get can  be shown by ferret (which is also a software
>>> package
>>> used by earth science communities) and Grads (Grads can read binary and
>>> data).
>>>
>>> The problem may occurred by conversion...
>>>
>>> Xueli
>>>
>>> Moreland, Kenneth wrote:
>>>      
>>>> I don't know what you mean by a "grad" or what problem it is causing,
>>>>so
>>>> I'm going to make a total guess.  I am guessing that you mean gradian
>>>>(a
>>>> unit equal to 9/10 of a degree).  Furthermore, because I can't think
>>>>of
>>>> any other way such units would mess you up, I am guessing that you are
>>>> trying to write spherical coordinates in terms of gradians, but
>>>>ParaView
>>>> is interpreting them either as degrees or as Cartesian coordinates.
>>>>
>>>>
>>>> Assuming that this is the problem, it is caused by the fact that the
>>>> default netCDF reader for ParaView honors the CF convention
>>>> (http://cf-pcmdi.llnl.gov), and the CF convention specifies spherical
>>>> coordinates in terms of degrees.  To specify angles in any other units
>>>> is
>>>> to break the convention.
>>>>
>>>> Your best solution is to convert your gradians to degrees (it's OK to
>>>> have
>>>> floating-point degrees) and write a netCDF file that properly follows
>>>> the
>>>> CF convention.
>>>>
>>>> If I have completely missed the point of your question, you'll have to
>>>> better explain what are grads and what problem you are having with
>>>>them.
>>>>
>>>> -Ken
>>>>
>>>> On 3/12/12 10:13 AM, "wang"<wang at knmi.nl>   wrote:
>>>>
>>>>
>>>>        
>>>>> Dear Paraview users,
>>>>>
>>>>> I have problem with data that origin of grads. We have a program to
>>>>> convert grads data into NetCDF format. But this data can not be shown
>>>>> in
>>>>> Paraview. I don't have any problem to show normal nc data.
>>>>>
>>>>> Have you ever experience the same problem?
>>>>>
>>>>> Thanks!
>>>>> Xueli
>>>>> _______________________________________________
>>>>> Powered by www.kitware.com
>>>>>
>>>>> Visit other Kitware open-source projects at
>>>>> http://www.kitware.com/opensource/opensource.html
>>>>>
>>>>> Please keep messages on-topic and check the ParaView Wiki at:
>>>>> http://paraview.org/Wiki/ParaView
>>>>>
>>>>> Follow this link to subscribe/unsubscribe:
>>>>> http://www.paraview.org/mailman/listinfo/paraview
>>>>>
>>>>>
>>>>>          
>>>>
>>>>        
>>>
>>>      
>>
>>    
>
>