[New data item]: Method for indicating the data source for a structure - experimental vs calculated

[rowlesmr]

Definition

It would be nice to be able to easily distinguish structures derived from experimental data vs calculations (eg DFT).

There is currently no way to do this in a machine-readble manner.

Restricted values

Value should be drawn from a predefined list

Example

experimental

Explanation of example

the data source for this structure is experimental, (as opposed to calculated)

Looping

top level

Data name

_diffrn.data_source

Type

Word (text with no spaces)

Data structure

None (a single value of type given in Type above)

Comments

_diffrn.data_source experimental
_diffrn.data_source calculated

Suggestions from meeting between CPD and IUCr.

[vaitkus]

Relevant discussion in the coreDMG mailing list: https://www.iucr.org/__data/iucr/lists/coredmg/msg00429.html

There was also a similar discussion by the OPTIMADE developers on introducing a similar enumerator (Materials-Consortia/OPTIMADE#455), however, it was quite difficult to agree on a limit where "experimental" ends and "theoretical" begins. The different types of "theoretical" methods also turned out quite difficult to agree upon.

[rowlesmr]

I missed _exptl.method in searching core.

That does look like the place to put this information, and yes, an enumeration does have a nice appeal.

[jamesrhester]

I think it would be good to pick up @vaitkus 's suggestions in the COMCIFS thread he references above.

The issue is basically whether or not to redefine our own _exptl.method to bring it into line with mmCIF - how many CIFs would we invalidate because they didn't have recognised enumerated values?