update PD_CALIBRATION

[rowlesmr]

EDIT: Lots of things have changed. Jump to the bottom for the current version.

.

Will close #133. Also applicable to #132.

This PR enhances PD_CALIBRATION by providing a category key based on the diffractogram_id to which the calibration applies, and a dataitem to record the diffractogram_id of the reference diffraction pattern from which the calibration was made.

[rowlesmr]

Question:
Both _pd_calibration.diffractogram_id and _pd_calibration.ref_diffractogram_id are defined as

    _name.linked_item_id          '_pd_diffractogram.id'
    _type.purpose                 Link

Does this mean that if neither of these is given a value, they take on the value of _pd_diffractogram.id if it is given in the same data block?

eg:

data_aDiffractionPattern
	_pd_diffractogram.id de14a39d-0d2b-44d5-b830-522b8c68f441 

	_pd_calibration.conversion_eqn          "E = 50000 + 100 * ch#"
	_pd_calibration.special_details         "Looks like the best fit"
	# _pd_calibration.diffractogram_id      takes the value of _pd_diffractogram.id from above?
	# _pd_calibration.ref_diffractogram_id  takes the value of _pd_diffractogram.id from above?

	loop_
	_pd_data.point_id
	_pd_meas.detector_id
	_pd_proc.energy_detection
	_pd_meas.counts_total
	a   0    50000    1234
	b   1    50100    1434
	c   2    50200    1457
	#...

The intent is that _pd_calibration.diffractogram_id is the diff_id of the diffraction pattern to which the calibration is being applied, and _pd_calibration.ref_diffractogram_id is the diff_id of the diffraction pattern from where the calibration was calibrated.

Would _pd_calibration.ref_diffractogram_id be better defined as

    # _name.linked_item_id          '_pd_diffractogram.id' #not this at all
    _type.purpose                 Encode

?

If the calibration were determined by some other method than a reference diffraction pattern, then I don't want to have a value for ref_diff_id. But, then again, a diffraction pattern to which a calibration is applied, then becomes a secondary reference pattern, so...

[rowlesmr]

What's above is the wrong way of doing it.

It should work much in the way PD_CALIB_XCOORD_OVERALL works.

It needs:

• _pd_calibration.id - key
• _pd_calibration.conversion_eqn - the equation to calibrate whatever is going on
• _pd_calibration.diffractogram_id - the diffractogram used to create the calibration

We can find the instrument and detectors from _pd_diffractogram.instr_id

This is basically formalising what it already was.

.

The next question is to whether the key should be changed.

.id - its human-readable, not machine-readble, so a person needs to figure out how to apply it. This is basically what already exists. It does require that you explicitly add an id value.

.diffractogram_id - this would be in lieu of .id. Its human-readable, not machine-readble, so a person needs to figure out how to apply the equation. Allows a lookup of _pd_diffractogram.instr_id to see in an instrument or detector_id has a calibration quation associated with it.

.detector_id - I think this is straying too far into the machine-readble phase with a thing that is fundamentally human-readable.

.something_else - anything else?