search

COMCIFS / Powder_Dictionary

CIF definitions for powder diffraction
4 stars 4 forks source link

why is it `_pd_proc.intensity_bkg_calc` and not `_pd_calc.intensity_bkg`? #72

Open rowlesmr opened 1 year ago

rowlesmr commented 1 year ago

The description of _pd_proc.intensity_bkg_calc says that it

is intended to contain the background intensity for every data point where the background function has been fitted or estimated (for example, in all Rietveld and profile fits).

PD_CALC says that it

is used for storing a computed diffractogram trace ... such as ... computed intensities from a Rietveld refinement.

Should there be a _pd_calc.intensity_bkg instead/as well as?

_pd_calc.intensity_* is of purpose Number, and _pd_proc.intensity_bkg_calc is of purpose Measurand (and "Inclusion of s.u.'s for intensity_bkg_calc values is strongly recommended.") , so there is a probably a difference of intent here between the two data items.

briantoby commented 1 year ago

Here is what I suspect was taking place, but I do not remember the details for CIF item naming/categorization: One collects data with a point spacing determined by the instrument, and one wants to keep these original as-collected data. However for the purposes of refinement, one may do some rebinning. Fewer points means faster fits, as long as there are enough points to not degrade resolution. Thus, one may have two sets of (x,y) points, original and rebinned. The background, calc pattern etc. will only be computed against the rebinned x-axis. Since there might be one loop or two for the data and computed patterns, the data names needed to be separated.

Brian

On Jan 7, 2023, at 10:30 AM, Matthew Rowles @.**@.>> wrote:

The description of _pd_proc.intensity_bkg_calc says that it

is intended to contain the background intensity for every data point where the background function has been fitted or estimated (for example, in all Rietveld and profile fits).

PD_CALC says that it

is used for storing a computed diffractogram trace ... such as ... computed intensities from a Rietveld refinement.

Should there be a _pd_calc.intensity_bkg instead/as well as?

_pdcalc.intensity* is of purpose Number, and _pd_proc.intensity_bkg_calc is of purpose Measurand (and "Inclusion of s.u.'s for intensity_bkg_calc values is strongly recommended.") , so there is a probably a difference of intent here between the two data items.

— Reply to this email directly, view it on GitHubhttps://github.com/COMCIFS/Powder_Dictionary/issues/72, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ACH7E2D4DS54NVPFPQ3LAFTWRGK3JANCNFSM6AAAAAATUBB66I. You are receiving this because you are subscribed to this thread.Message ID: @.***>

rowlesmr commented 1 year ago

mmm. There must have been some context and reasoning for the split.

The _pd_calc.intensity_* data items say that "values should be computed at the same locations as the processed diffractogram", so a _pd_calc.intensity_bkg item seems to be a drop in replacement for _pd_proc.intensity_bkg_calc, except for the "estimated" part of the definition.