search

materialsvirtuallab / maml

Python for Materials Machine Learning, Materials Descriptors, Machine Learning Force Fields, Deep Learning, etc.
BSD 3-Clause "New" or "Revised" License
374 stars 79 forks source link

Installing GAP and NNP dependencies #166

Closed vishankkumar closed 3 years ago

vishankkumar commented 3 years ago

Hi,

I was trying to run the example of PES with GAP fitting, and would like to know how I can interface MAML code with GAP. I have installed quippy with GAP already but do not know how to let maml code direct at the correct file/location to look for GAP capabilities.

Thanks

YunxingZuo commented 3 years ago

Hi @vishankkumar

Thanks for your interest in using PES module!

Please add the capabilities (e.g., gap_fit for GAP fitting) in your environment path. The fit method works by detecting the executable automatically and run it in the background. If the fit method doesn't detect the executable, it will throw an error of "gap_fit has not been found."

vishankkumar commented 3 years ago

Dear @YunxingZuo ,

Thank you for your response. I managed to get the Gap_fit executable linked properly, however I do not know how to link nnp train from n2p2. I have compiled nnp-train in the Shared mode.

Thanks for your response. Vishank

vishankkumar commented 3 years ago

I got it worked, I just had to add the lib/ to LD_LIBRARY_PATH. However, I am getting an error in the example notebooks/pes/nnp/. I have attached the snapshot of the error. Would you please let me know what is the reason for that. screenshot_NNP_Error

YunxingZuo commented 3 years ago

@vishankkumar Could you please update the maml to the newest version (i.e., 0.0.4)? I believe this problem has been solved in #194.

vishankkumar commented 3 years ago

@YunxingZuo Thanks for the suggestion, it worked after pulling the latest version from github.