Universite-Gustave-Eiffel / I-Simpa

An Open Source software for 3D sound propagation modelling
https://i-simpa.univ-gustave-eiffel.fr/
GNU General Public License v3.0
227 stars 56 forks source link

Export of results into hdf5 #278

Open qmichalski opened 3 years ago

qmichalski commented 3 years ago

Hi,

First, I want to state how really helpful this software has been to the project I am working on.

I am looking at exporting the field results to be able to process them further separately for instance in Python, I couldn't find in the documentation how the type "proj" or the "rs_cut.csbin" is structured to be able to access it.

Would the documentation be amended at some point to give a better insight on how the data is saved, would the export of field data be something you would consider implementing in the future?

Kind regards,

Q.M.

[Edit] I checked again the doc, there are things, but the descriptions are limited. For instance for the surface receiver, what term the output of GetFaceSumEnergy refers to?
[Edit2] Ok so i spent some time looking at some work in progress and UI function and found the conversion function. I can upload a version of the Python script I have where I convert the receptor to .vtk so that it can be read in Paraview. The scene (wall et al.) can then be exported as .ply from I-SIMPA to be then directly read into Paraview.

nicolas-f commented 3 years ago

Hi,

Thanks for your interest in I-Simpa. Surface/Cut receivers results should not be used as they are not covered by validation. You should export results of punctual receivers instead.

best regards,

Nicolas

qmichalski commented 3 years ago

Hi Nicolas, thanks for your answer!

Yeah, I figured I would only use the punctual receivers for quantitative comparisons. Out of curiosity, I compared local points sampled by punctual receivers to points taken from the field. The only discrepencies i found where when the gradients in out-of-plane direction were high. I assumed it is because the receptor is sampling as a volume and not a surface. All of that was done making sure that the sampling size for the surface receptor was the same as for the punctual receivers. Do you guys plan to validate those further at any point? I imagine it's a hard one given that experimentally there is no such thing as "surface microphone".

Again thanks for your time, thanks for the great software and libraries !

Cheers,

Q.M.

qmichalski commented 1 year ago

Hi Nicolas,

Someone contacted me about it, so I reckon it could be useful to others. As i said above, i understand and acknowledge that you guys have not validated the surface receiver. That said, those receivers do give a good qualitative representation of what's going on and it can be good to be able to export that to appropriate format to do something with it. For those interested I made a small script using your existing examples to convert the content of surface receivers to vtk so that you can import this into data rendering software like Paraview, etc. It's sitting there for now, if you want to do anything with it, https://github.com/qmichalski/i-simpa-vtk.

Cheers,

Quentin

nicolas-f commented 1 year ago

Hi,

Thank you for sharing your work. Maybe we could add a link to your repository in the documentation.

Best regards,

-Nicolas