[Paraview] [EXTERNAL] "Select Cells On" large object?

[Paul McIntosh]

Hi Ken,

Thanks for the suggestion - I've done some experiments and I think the best
way to do it is extract sections close up to get the resolution and then
postprocess the bits back together. I tried a Python trace hoping that I
could automate the process, however the actual selection region was not part
of the trace (just the list of cells it extracts).

Also another issue is that extract selection does not maintain the cell
id's, so the data gets doubled up in the overlapping regions if it is simply
appended the resulting extracted sets. The solution seems to be to use the
Python trace and merge all the selection id's into one call, then the same
cells are treated as such. It seems to work with the sphere, I'll try it
next on my data. 

Cheers,

Paul

-----Original Message-----
From: Moreland, Kenneth [mailto:kmorel at sandia.gov] 
Sent: Thursday, 15 December 2011 2:50 AM
To: Paul McIntosh; paraview at paraview.org
Subject: Re: [EXTERNAL] [Paraview] "Select Cells On" large object?

Short answer, no.

If your surface is convex, you could threshold the surface elements you need
by first running the extract surface filter, then generating cell normals,
then using the calculator to extract the x component of the cell normals,
then thresholding all cells with the x component of the normal greater than
0.  The remaining steps are the same as those you sent from the transform
filter on.

If your surfaces are not convex, perhaps you should consider an image-based
computation.  Set up the camera as in the instructions you sent.  Then
figure out the cross-sectional area of the screen and multiply that by the
ratio of pixels covered by the object to the total number of pixels.  This
ratio would probably be easiest computed if the object color was set to
black and the background to anything else.  Perhaps one of the Python gurus
could give some hints as to how to automate some of this.

-Ken



On 12/13/11 6:37 PM, "Paul McIntosh" <paul.mcintosh at internetscooter.com>
wrote:

>Hi,
>
>Hopefully an easy question to answer...
>
>"Select Cells On" is "visible" resolution dependent. By this I mean, 
>for the same selection area, if I am close to the surface I get all the 
>cells selected but if I am far away many cells are missed.
>
>I am calculating frontal area of a large object where I cannot get the 
>whole object selected and get close enough to see all the individual 
>cells. The result is missing cells which are mucking up my calculation. 
>Is there an easy way to get all cells selected on a large object?
>
>Note: I am using this method for working out frontal area:
>http://www.discretizer.org/node/20
>
>Cheers,
>
>Paul
>---
>www.internetscooter.com
>
>_______________________________________________
>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
>