Trying to follow compressed sensing notebook, but I get stuck since [atomic_properties_lda2015.symbol(el).get('atomic_'+f) for f in features] does not cover many elements e.g. 'Au','Co','Cr','Fe','Hf','Ir','Mn','Mo'. I see there are also other functions to retrieve atomic properties like atomic_properties_pymat which takes in the above elements, but I dont know how the features are named so I can list them.
Is there a documentation to this? how do I access these from jupyter notebook ?
Thanks for noticing this.
Those atomic properties were created specifically for that dataset. If you wish to add more atomic properties I suggest that you use the example shown in the query_nomad_archive.ipynb notebook.
https://nomad-lab.eu/prod/analytics/public/user-redirect/notebooks/tutorials/query_nomad_archive.ipynb
Thanks!
I just realized that that the notebook I worked on was completely overwritten with the original compressed_sensing.ipynb. Do I have to save the notebooks manually ? Or is it saved somewhere?
Notebooks in the tutorials directory are reinitialized at any new log in. You can definitely make modifications and test different features, but, once your server is stopped and then you log in again, those modifications are lost. If you wish to create permanent changes, you should use the work directory, where saved modifications are reaccessible. So the best solution is that you copy the compressed sensing tutorial in the work directory and then work on that copy.
Please note that a newer version of the toolkit is in production, where the latest atomicfeaturespackage version is available. Differences are minimal, and you can take a look at query_nomad_archive and compressed_sensing tutorials where the package is deployed.
Thanks.
Is there documentation for the atomicfeaturespackage? the notebooks you propose do not show all the properties that I can call from this package.
Sure!
You can find a preliminary documentation here:
As soon as documentation is completed a link in the tutorials will be provided.