search

COMCIFS / Powder_Dictionary

CIF definitions for powder diffraction
4 stars 4 forks source link

Created PD_CALIB_WAVELENGTH #91

Closed rowlesmr closed 1 year ago

rowlesmr commented 1 year ago

This category holds information on the diffractograms and phases used to calibrate the incident wavelength.

jamesrhester commented 1 year ago

So I think that the phase_name is superfluous as that will be available via the phase_id anyway.

There needs to be a way to specify which wavelength was calibrated, as some data sets may combine data from multiple instruments or instrument settings. These wavelengths are keyed to _diffrn.id, so logically there should be a data name in PD_CALIB_WAVELENGTH that is a child of _diffrn.id. I realise that diffraction conditions can change without changing the wavelength but that is a sub-optimal cif_core design decision that we can't change.

I don't think we need _pd_calib_wavelength.id as _phase_id, _diffractogram_id and _diffrn.id together uniquely identify a particular phase in a particular diffractogram used to calibrate wavelength under a particular set of diffraction conditions. Or are we considering multiple measurements of the same phase in order to calibrate wavelength? I think that would be going too far.

I think the wavelength calibration story is now pretty good. To include wavelength calibration information, simply include:

  1. A phase-specific data block containing the cell parameters and name of the standard material
  2. A diffractogram from that material
  3. PD_CALIB_WAVELENGTH information in the data block holding overall information per set of diffraction conditions (single value of _diffrn.id).

This is enough information for software to automatically rederive the calibrated wavelength, which is our ultimate goal.

rowlesmr commented 1 year ago

So I think that the phase_name is superfluous as that will be available via the phase_id anyway.

It's gone. (I'll also remove other instances of _*.phase_name that are like this (in other PRs))

I don't think we need _pd_calib_wavelength.id as _phase_id, _diffractogram_id and _diffrn.id together uniquely identify a particular phase in a particular diffractogram used to calibrate wavelength under a particular set of diffraction conditions. Or are we considering multiple measurements of the same phase in order to calibrate wavelength? I think that would be going too far.

Doesn't the combination of _phase_id, _diffractogram_id and _diffrn.id allow for any number of phases and diffractograms to contibute to the one particular set of diffraction conditions? Multiple measurements of the same phase would give different diffractograms, many phases would have different phase_ids with the same diffractogram id.

jamesrhester commented 1 year ago

Doesn't the combination of _phase_id, _diffractogram_id and _diffrn.id allow for any number of phases and diffractograms to contibute to the one particular set of diffraction conditions? Multiple measurements of the same phase would give different diffractograms, many phases would have different phase_ids with the same diffractogram id.

Yes, you are right. So even multiple measurements of the same phase under the same diffraction conditions would yield a unique row in this category, and so the _pd_calib_wavelength.id data name is even less needed.

jamesrhester commented 1 year ago

Thanks for all the work!