mfraters / cpo_analyzer Goto Github PK

Currently only providing olivine or enstatite input data doesn't work. If you have both data you can plot just one, but you need both data sets. This should be fixed.

• allow having just olivine data (priority)

• allow having just enstatite data

in #21 I was not able to fix the macos tester which worked before on the exact same code and now doesn't anymore. I tried to use macos-12, but that also didn't help. Disabled it for now until a solution can be found.

Currently the tester outputs in the experiment directory. I think this done to keep the code simple (output where the data is), but it is not a clean setup. A solution would be to allow for a different output directory, or let the tester first copy the files to a more suitable location and then run it.

The idea is to get the CPO averaged within a certain distance of a point or of grid points. This would allow for visualization on map to get a sense of what the anisotropy would look like. For a better approach the data would need to be processed by something like ECOMAN, but this might be good enough for now.

Hello,

I am independently working on some CPO modelling (which may eventually be superseded by ASPECT with your well-written plugin) and I'd like to try using this package for visualisation. I'm not clear on the input format, because I'm not using ASPECT. I tried to run the shearbox models from your 2021 paper, however I had trouble getting aspect to run the sample input file provided in the supplementary material:

----------------------------------------------------
Exception 'ExcMessage("Value \""+ write_draw_volume_weighted_lpo_list[i] +"\", set in \"Write out draw volume weighted lpo data\", is not a correct option " + "because it should contain a mineral identification and a output specifier seprated by a colon (:). This entry " + "does not follow those rules.")' on rank 0 on processing:

--------------------------------------------------------
An error occurred in line <955> of file </home/admin/srcbuilds/ASPECT-Fraters/source/postprocess/particle_lpo.cc> in function
    void aspect::Postprocess::LPO<dim>::parse_parameters(dealii::ParameterHandler&) [with int dim = 3]
The violated condition was:
    split_draw_volume_weighted_lpo_instructions.size() == 2
Additional information:
    Value "olivine Euler angles", set in "Write out draw volume weighted
    lpo data", is not a correct option because it should contain a mineral
    identification and a output specifier seprated by a colon (:). This
    entry does not follow those rules.
--------------------------------------------------------

Aborting!
----------------------------------------------------

I will try to fiddle with the weighted output parameter names. The relevant section of the parameter file is:

    subsection LPO
        set Time between data output = 1e5
        set Write in background thread = true
        set Compress lpo data files = true
        set Write out raw lpo data =
        #set Write out draw volume weighted lpo data = mineral 0: volume fraction, mineral 0: Euler angles, mineral 1: volume fraction, mineral 1: Euler angles
    end

So really I'm sneaking a few questions in here:

• Do you perhaps have an up-to-date parameter file for the aspect CPO plugin? I'm running from the CPO_afterpaper_changes branch.
• In the example folder, it seems that the input for this program is Euler angles, stored in separate .dat files for each grain. Is that correct? Maybe it would be convenient to allow input of multiple grain data from one file? Are there other ASPECT-specific things in the input that I should be aware of, or can I just turn off compression and match the .dat format? I will try a few things, unfortunately I'm not familiar with Rust but maybe I can help a little bit.

Thanks for making your code and methods public and accessible, it has already helped me a lot.