[Paraview] [Paraview-developers] The Quest for Data

[Eric E. Monson]

Thanks, Utkarsh -- I appreciate it.

With the way the existing query-based selection framework is implemented, is there any chance of incorporating the chaining of queries to select subsets of subsets? (With my users, I know this is one of the first things they're going to ask me when I show them the new functionality.)

I can add these as feature requests, but it would also be cool if there was a way to specify a precision for some types of selections, like within a certain error tolerance of a floating point value, or within a certain radius of a point coordinate. (I know the "is between" does the former, but for some applications the value +/- tolerance might be more natural to think about.)

Talk to you later,
-Eric


On Mar 1, 2010, at 2:42 PM, Utkarsh Ayachit wrote:

> Eric,
> 
> Good catch. I must have missed this one. The problem seems to be that
> vtkExtractSelectedIds has a stupid limitation where it expects the
> array types to be same. I am going to look at fixing
> vtkExtractSelectedIds soon (give me a day or two).
> 
> Utkarsh
> 
> On Mon, Mar 1, 2010 at 2:07 PM, Eric E. Monson <emonson at cs.duke.edu> wrote:
>> Hey Utkarsh,
>> 
>> That's very cool functionality.
>> 
>> One part of the interface is either not very intuitive or I'm running into a bug: The "is" or "is one of" operators don't seem to work with anything other than ID or Global ID. If I try to use it, say, on an integer point data attribute and do "attribute is 0" I get this:
>> 
>> Warning: In /Users/emonson/Programming/ParaView_cvs/ParaView3/VTK/Graphics/vtkExtractSelectedIds.cxx, line 666
>> vtkExtractSelectedIds (0x12c5f50e0): array types don't match
>> 
>> ERROR: In /Users/emonson/Programming/ParaView_cvs/ParaView3/VTK/Filtering/vtkExecutive.cxx, line 757
>> vtkStreamingDemandDrivenPipeline (0x11f3d63b0): Algorithm vtkExtractSelectedIds(0x12c5f50e0) returned failure for request: vtkInformation (0x12bd2f490)
>> Debug: Off
>> Modified Time: 2052406
>> Reference Count: 1
>> Registered Events: (none)
>> Request: REQUEST_DATA
>> FORWARD_DIRECTION: 0
>> FROM_OUTPUT_PORT: 0
>> ALGORITHM_AFTER_FORWARD: 1
>> 
>> I get the same thing when I try to do that type of query on a float or idtype attribute, too.
>> 
>> So, is "is" supposed to work with integer and other types like the "==" operator, or just with ID and Global ID?
>> 
>> Thanks,
>> -Eric
>> 
>> ------------------------------------------------------
>> Eric E Monson
>> Duke Visualization Technology Group
>> 
>> 
>> On Feb 24, 2010, at 4:14 PM, Utkarsh Ayachit wrote:
>> 
>>> Folks,
>>> 
>>> How many times have we wondered:
>>> * Where's the cell with DISPLACEMENT(Magnitude) == 12?
>>> * What's the point where the error condition is reached?
>>> * What cell is at the location (1,2 3)?
>>> 
>>> We always knew selection mechanism in ParaView could do that, but we
>>> never found how to get it done -- the selection inspector was just too
>>> confusing. Fret no more: we've added new functionality to ParaView to
>>> help you locate the cells/points of interest.
>>> 
>>> Refer to following Wiki page for details.
>>> 
>>> http://www.paraview.org/Wiki/Find_Data_using_Queries
>>> 
>>> This is currently in CVS (very new, so bug reports/feature requests
>>> are always welcome) and should make it into the forthcoming 3.8
>>> release.
>>> 
>>> Utkarsh
>>> <Find Data.png>_______________________________________________
>>> Paraview-developers mailing list
>>> Paraview-developers at paraview.org
>>> http://public.kitware.com/mailman/listinfo/paraview-developers
>> 
>>