Closed jamesrhester closed 1 year ago
Just a note on the interaction between _pd_block.diffractogram_id
and pd_meas/proc
. _pd_block.diffractogram_id
should point to data blocks containing a diffractogram. It is possible that this block could contain both a pd_meas
loop of raw(ish) data, and a pd_proc
loop of processed data (see example in IT Vol G 3.3.7.1). So both pd_meas
and pd_proc
should contain children of a notional pd_diffractogram.id
.
Dependence of a piece of data on a particular diffractogram would normally be expressed in CIF by adding a child of some diffractogram identifier to the category containing that data. Currently pdCIF simply uses the block identifier as a proxy for the diffractogram identifier, with a system of block pointers linking particular data to particular diffractograms. There is no need to change this system. By adding a
PD_DIFFRACTOGRAM
category with identifier (and that identifier being made equivalent to the block_id) it will be possible to express the relationships between diffractograms and data in a way that is robust, plays well with other dictionaries, and is mathematically rigorous, while not changing the current "pointer" systemThis will also allow various schemes for distributing powder data over data blocks to be formally described in terms of categories, with
PD_DIFFRACTOGRAM
corresponding to diffraction data just asPD_PHASE
corresponds to phase.