[Paraview] Unstructured Grid Descriptive Statistics

[David Thompson]

Hi Kristjan,

You are correct that the statistics filters do not weight values by  
cell area. It's torturous, but it might be possible to use the "Mesh  
Quality" filter to compute the cell areas/volumes, a calculator filter  
to compute the product of the area and a cell variable, and then run  
descriptive statistics on the resulting field. But being able to  
perform area/volume-weighted statistics (at least for cell data)  
sounds like a good use case to support. Would you be willing to file a  
feature request at http://paraview.org/Bug/ and attach your example to  
it?

	Thanks,
	David

On Dec 7, 2010, at 7:31 , Kristjan Onu wrote:

> Hi,
>
> I'm having difficulty using the Descriptive Statistics filter with  
> data
> specified on unstructured grids.
>
> The two files attached to this email demonstrate the problem. The mean
> of y obtained with 2_cell_square.vtu is 0.5 whereas with
> 5_cell_square.vtu the mean is 0.417.
>
> I think this happens because the mean is calculated over all the  
> points
> without taking into account the cell area. Can anyone explain what  
> steps
> I should follow to obtain a mean of 0.5 with 5_cell_square.vtu?
>
> Thanks,
>
> Kristjan
> <2_cell_square.vtu><5_cell_square.vtu><ATT00001..txt>