search

openPMD / openPMD-example-datasets

:gift: HDF5 Example Files
Creative Commons Zero v1.0 Universal
8 stars 4 forks source link

Close #4 and #5: no particle in 3D example & erroneous basePath #6

Closed RemiLehe closed 8 years ago

RemiLehe commented 8 years ago

This pull requests loads new datasets:

For documentation and reproducibility, I also added the scripts that produce these datasets.

PrometheusPi commented 8 years ago

Thanks @RemiLehe. I will have a look at the new data sets tomorrow.

PrometheusPi commented 8 years ago

Great job @RemiLehe. If it is fine for you and @ax3l that the example-thetaMode has no particles, this pull request can be merged. (I am no maintainer, so I can not merge.)

ax3l commented 8 years ago

@RemiLehe thank you a lot for the quick update and thanks to @PrometheusPi for testing!

ax3l commented 8 years ago

yes if @RemiLehe agrees we can also add particles in the theaMode sims.

PrometheusPi commented 8 years ago

Warning: You added one zero to the file names. Thus old scripts reading the data will not work.

RemiLehe commented 8 years ago

@PrometheusPi Regarding your remarks on the number of zeros, I was not sure if you are asking me to remove that zero, or if it was a personal note for yourself to remember to adapt your old scripts. Can you clarify ?

Regarding the particles for thetaMode, if it is okay with you I would still leave the file without particles, as it used in the openPMD-viewer tutorial, in order to show that the viewer can adapt to the absence of particles, in the GUI. But it is definitely not a very important point, so if you actually need the particles in the data of thetaMode, I can definitely add it. Please let me know.

ax3l commented 8 years ago

I think he means your "general scripts that read those files", e.g., our test scripts in openPMD-viewer? Also be aware that I just realized https://github.com/openPMD/openPMD-standard/issues/121 today for detecting the iteration from the file name.

So right now, using file names of the form data_100.h5 and data_50.h5 will likely be easier for reader implementations than padding zeros such as data_00100.h5. We can fix/refine this in 1.X.

ax3l commented 8 years ago

After thinking once more about: it is not that much of a problem and we should just be more explicit that this can be zero-padded.

RemiLehe commented 8 years ago

As far as I can tell, openPMD-viewer works fine with any number of zeros. (and reads the datasets of the PR correctly) @ax3l If I understand you, I can leave the filenames as they are, is that correct ? If yes, you can probably merge the PR.

ax3l commented 8 years ago

absolutely, thank you a lot! :sparkles: