Every parameter that can be refined in GSAS-II is defined a unique name that is associated with a value. These names are referenced when defining constraints and restraints. The naming follows this convention:
P:H:name
where P is a phase number and H is a histogram number and name specifies the type of parameter. The name values are tabulated in the GSAS-II Developer’s Documentation in the table at this URL: https://gsas-ii.readthedocs.io/en/latest/objvarorg.html∖#parameter-names-in-gsas-ii. When not appropriate, H or P can be blank. Global parameters, which are very uncommon, do not have an H or P value. Note that internally to GSAS-II, parameter names are stored in a fashion that abstracts the phase and histogram numbers, so that if the phases or histograms are reordered, the names will automatically be updated to follow.
As an example of how a number would be omitted, a wavelength is a parameter of a histogram and not a phase and the name LAM is associated with wavelength. Thus, the wavelength for histogram 3 will be named :3:LAM .
Atom parameters are named with an additional number (note that the histogram number is omitted as atoms are not associated with phases:
P::name:atom-num
As an example of that, 1:Uiso:12 is the \(\rm U_{iso}\) value for atom 12 in phase 1. Note one complication for atom coordinates. While atom coordinates parameter names for atom n in phase p are p:Ax:n, p:Ay:n, and p:Az:n, the parameters that are varied are named p:dAx:n, p:dAy:n, and p:dAz:n. These “dA*” values are offsets that are applied to coordinates rather than the coordinates themselves. In this way when an atom is at a site x,1/2-x,z only the change, \(\delta \)x is computed and is added to the x coordinate and subtracted from the y coordinate.
Finally, as noted in §14.7, the HAP parameters have values for each phase and each histogram. As an example there can be March-Dollase texture parameter for every phase and every histogram and it will be named 1:2:MD for phase 1 and histogram 2.
Alas, there is one place where this naming is confusing, but a change would likley introduce bugs that we would need to spend years finding and fixing. Every histogram has a scale factor, named :0:Scale (etc.), but there are HAP parameters for phase fractions and these use the same name, 0:0:Scale (etc.). They are distinguished by the scale factor having no phase number while the phase fraction has both.