Conflicts between theoretical tag and icsd_ids tag

I realized if an entry has matched structures in ICSD, it will be regarded as an experimentally discovered phase, then tagged as theoretical=False, as this discussion in the forum. However, I found many entries that contain icsd_ids records are tagged with theoretical=True in the database V2021.03.22, such as mp-1192350, mp-1080443, mp-14208, etc.

I should trust which tags to determine an experimentally discovered entry.


Hi @Peking_jo

It seems I wasn’t very clear in that post. It’s a bit more complicated than that. The ICSD has both theoretical and experimental structures and tags them as such. So if the ICSD ID is tagged as experimental, we will too. If all the ICSD IDs are tagged as theoretical by the ICSD, then we keep it as a theoretical compound.

Theoretical compounds are tagged that only as far as we know. We’re working on integrating in more databases such as COD and Pauling Files.

My question appeared when I ran the codes:

>>> entries = mpr.query(criteria={'material_id':{'$in':['mp-14208','mp-1080443','mp-1192350','mp-2116']}}, properties=['theoretical', 'icsd_ids'])

properties were returned as

>>> [{'theoretical': True, 'icsd_ids': [185943]}, {'theoretical': True, 'icsd_ids': [188263]}, {'theoretical': True, 'icsd_ids': [262362]}, {'theoretical': True, 'icsd_ids': [246489, 246490, 246491]}]

Now, I see, those ICSD entries are collected from ab initio calculation papers, and may be tagged as theoretical structures in the ICSD, then the attributes are carried on in MP database. Thank you for your interpretation.