Predicted elastic moduli in new MPRester

Materials Project
1

Hi all,

I am switching to the new MPRester which gets the data from https://next-gen.materialsproject.org and so far it worked like a charm. But I am wondering how to get predictive elastic moduli with the new rester. In the old one I could do

MPRester.get_data(, data_type=“pred”, prop=“elastic_moduli”)

but get_data is not available in the new rester.

Best,

Josua

2

Hi @josua_exomatter,

We haven’t carried this forward because the method is very old right now; there are much improved methods available from other groups, e.g. MEGNet to mention just one.

However, these predicted moduli can still be generated directly via the gbml code, some amount of hacking may be required but it’s open source and fairly simple.

Hope this helps,

Matt

3

As Matthew suggested me, I saw that there’s an API for MegNET.

I tried to use it in this way and it gave me the same results reported in their website (http://megnet.crystals.ai/):

from pymatgen.ext.crystalsai import CrystalAIRester
CrystalAIRester().predict_mp('log10K','mp-1143')

see also CrystalAIRester().get_available_models() and docs of predict_mp and predict_structure funcs.