search

sparks-baird / CrabNet

Predict materials properties using only the composition information!
https://crabnet.readthedocs.io/
MIT License
12 stars 5 forks source link

FileNotFoundError: [Errno 2] when calling crabnet.utils.figures.element_prevalence #40

Open Pepe-Marquez opened 2 years ago

Pepe-Marquez commented 2 years ago

First of all, I love the package, amazing work.

Just one tiny thing. When calling the element_prevalence utility I get the following error. It is not important, but I guess it is also no effort reporting it.

FileNotFoundError: [Errno 2] No such file or directory: 'ML_figures/element_properties/ptable.csv'

sgbaird commented 2 years ago

Thanks @Pepe-Marquez! Taylor and I are at a conference right now, and we both appreciated your shout-out on Twitter. Glad to hear that you like the package, and main kudos goes to Kaai Kauwe and Anthony Wang (former Sparks group members) for developing the main repo; I've just been refactoring, extending, and packaging it. Thanks for letting me know about the bug with the sub-utility. Will check into it. Interested to hear what use-cases you find for CrabNet in your research!

Pepe-Marquez commented 2 years ago

Thank you, Sterling! Unfortunately, I did not find Anthony nor Kaai on Twitter but I am well aware and I follow their work.

As I mentioned, the bug is really not important and comes if you pip to install the package. It can be overcome just but putting the files manually in the path that is hardcoded in the function. The same happens with the generate_features which rely on grabbing the tables from the data folder.

Regarding my use-case... I am only getting introduced to ML, I am a solar energy materials scientist. But now I work at Nomad-lab in the FAIRmat project. And I wanted to do something with it within this context. Nomad is changing very fast these days with FAIRmat and turning into a platform to cope easily with experimental data, so I think that having an example running on how to use CrabNet would be top. I have seen you commented at some point in the Nomad-lab forum, are you familiarised with it? Might be worth passing this conversation to proper emails :D.

Thank you!

sgbaird commented 2 years ago

Thank you, Sterling! Unfortunately, I did not find Anthony nor Kaai on Twitter but I am well aware and I follow their work.

All good!

As I mentioned, the bug is really not important and comes if you pip to install the package. It can be overcome just but putting the files manually in the path that is hardcoded in the function. The same happens with the generate_features which rely on grabbing the tables from the data folder.

Thanks for the clarification!

Regarding my use-case... I am only getting introduced to ML, I am a solar energy materials scientist. But now I work at Nomad-lab in the FAIRmat project. And I wanted to do something with it within this context. Nomad is changing very fast these days with FAIRmat and turning into a platform to cope easily with experimental data, so I think that having an example running on how to use CrabNet would be top. I have seen you commented at some point in the Nomad-lab forum, are you familiarised with it? Might be worth passing this conversation to proper emails :D.

If you're referring to Novel Materials Discovery (NOMAD), then yes, I'm a big fan! You may be interested in https://doi.org/10.6084/m9.figshare.19319783.v3

I think it's pretty relevant if you're interested in CrabNet. Maybe you already came across it.

A CrabNet discussions post (https://github.com/sparks-baird/CrabNet/discussions), email, or slack is fine by me. For the latter two: sterling.baird@utah.edu

Looking forward to chatting more! 🙂