[Paraview] Vis template by changing reader file names?

[Eric E. Monson]

Hey All,

My group wants to use ParaView for debugging their simulation system,  
where they might output many different files in a day with tweaked  
parameters and want to quickly view the results to see whether things  
are changing the way they expect.

The problem is that for each run there are three different output  
(unstructured grid) files, and to make a nice visualization each has  
to be glyphed or colored or filtered in some special way. I can do  
this, and it looks great, but it takes a while to set up each time a  
new set of files is generated, so it's a bit of a pain.

Therefore, main question: Once a visualization pipeline is set up, is  
there any easy way to pick new files for the readers so new data could  
quickly be visualized using an existing "template"?

I have tried reading in new files and using the "Change Input..." menu  
item, but it's a bit of a pain because it's a multi-step method that  
changes the order of items in the Pipeline Browser each time you do  
it. I know how to save a state file and then edit the source file  
names for the readers, but this takes the user out of ParaView and is  
a little clunky, but do-able. I can also write a python script that  
loads a state file and changes the input files and creates an  
animation automatically, but that isn't as flexible as looking at the  
data in ParaView itself. (I also tried to change a source file name  
from the Python shell within ParaView, but at least with an Xdmf file,  
that always crashes ParaView when I'm driving...)

Am I missing something clean and easy? Is this a typical thing people  
would want to routinely do with ParaView?

Thanks,
-Eric

------------------------------------------------------
Eric E Monson
Duke Visualization Technology Group