Correct keys for `PD_CALIB*` categories

[rowlesmr]

Writing as I go through adding things to Powder chapter so I can come back to them later.

Thoughts:

Should the calib categories be keyed essentially solely on _pd_instr.id\_pd_instr_detector.id?

PD_CALIB_D_TO_TOF I'm leaving alone, as TOF instruments are complicated

PD_CALIB_WAVELENGTH is currently keyed on *.diffractogram_id, *.diffrn_id, and *.phase_id. I think it should be *.diffractogram_id and _pd_instr.id; you're using a standard material to figure out the wavelgnth of the instrument, much like PD_QPA_EXTERNAL_STD.

PD_CALIB_XCOORD is currently keyed on *.diffractogram_id, and its own *.id. Should _pd_instr_detector.id be included, given PD_CALIB_XCOORD_OVERALL is keyed on _pd_instr_detector.id.

PD_CALIBRATION has no key.

PD_CALIB_INCIDENT_INTENSITY is keyed on _pd_instr.id
PD_CALIB_DETECTED_INTENSITY is keyed on _pd_instr_detector.id
PD_CALIB_XCOORD_OVERALL is keyed on _pd_instr_detector.id

[rowlesmr]

PD_CALIB_WAVELENGTH is not right. I think that this is like PD_QPA_EXTERNAL_STD, inasmuch as we analyse a diffraction pattern to get a value to calibrate an instrument. - see also #188

To this end, I suggest:

• PD_CALIB_WAVELENGTH should be a Set Category
• _pd_calib_wavelength.diffractogram_id is a key, linked to _pd_diffractogram.id of the diffractogram used to calibrate.
• _pd_calib_wavelength.instr_id is a key, linked to _pd_instr.id of the instrument being calibrated.
• _pd_calib_wavelength.radiation_id is the "result" of the calibration, linked to _diffrn_radiation.id

[rowlesmr]

from #133
PD_CALIBRATION should have a key _pd_calibration.diffractogram_id , linked to _pd_diffractogram.id of the diffractogram used to create the calibration equation. This mirrors the previous delegation of _pd_calib_std.external_block_id to denote the block containing the data whence the calibration came.

We can't have a _pd_instr.id key, as this calibration is about the detector side of things, and it can't have a _pd_instr_detector.id key, as one of the use-cases is to provide a conversion from detector_id to 2Th\d\q\whatever.

Does it need anything else? see #134

[jamesrhester]

Agree that PCW should be a Set category as all the current keys are Set category keys.

PCW.instr_id should not be a key because that is derivable already if you know the diffractogram.
PCW.radiation_id, similarly, is derivable if you know the diffractogram. So only one key is necessary.

Alternatively, we could completely drop PCW and simply add two columns to diffrn_radiation_wavelength, one giving the diffractogram and one giving the phase. How does that sound?

(A possible problem with the category is that it doesn't appear to explicitly link to the wavelength that was calibrated. If I want to start from a diffractogram to find which diffractogram was used to calibrate the wavelength by following links, I can't find my way to PCW. I can get to the actual wavelength by _pd_diffractogram.diffrn_id -> _diffrn.diffrn_radiation_id -> _diffrn_radiation_wavelength.radiation_id, which will give me a list of wavelengths. I can do the same as the above with PCW.diffractogram_id, so I guess that is implying that the wavelengths I've reached are those that were calibrated with this diffractogram.)

[rowlesmr]

Alternatively, we could completely drop PCW and simply add two columns to diffrn_radiation_wavelength, one giving the diffractogram and one giving the phase. How does that sound?

Yes, lets go with that!

[rowlesmr]

Further development here will probably depend on the outcome of #231