Sampled correlations #104
Sampled correlations #104
Conversation
docs/src/structure-factor.md
Outdated
| # Make a `StructureFactor` and calculate an initial sample | ||
| sf = DynamicStructureFactor(sys; Δt=0.05, ωmax=10.0, nω=100) | ||
| # Make a `SampledCorrelations` and calculate an initial sample | ||
| sc = SampledCorrelations(sys; Δt=0.05, ωmax=10.0, nω=100) |
There was a problem hiding this comment.
This no longer has an initial sample as suggested by the comment
docs/src/structure-factor.md
Outdated
| @@ -122,7 +115,7 @@ configurations. Information about calculating instantaneous data from a dynamic | |||
| structure factor can be found in the following section. | |||
|
|
|||
| The basic usage for the instantaneous case is very similar to the dynamic case, | |||
| except one calls `InstantStructureFactor` instead of `DynamicStructureFactor`. | |||
| except one calls `InstantStructureFactor` instead of `SampledCorrelations`. | |||
There was a problem hiding this comment.
Unsure what to do about InstantStructureFactor
There was a problem hiding this comment.
Seems weird to me that calling InstantStructureFactor builds you a SampledCorrelations (none of those words are the same anymore)
There was a problem hiding this comment.
Agreed. I thought this was just where we left it in the thread yesterday.
I guess we could have kept the whole Instant vs Dynamic dichotomy: SampledDynamicCorrelations and SampledInstantCorrelations.
There was a problem hiding this comment.
How about SampledCorrelations and InstantCorrelations, where "sampling" refers to the dynamical sampling process? I suppose there is the possibility that we might want to sample InstantCorrelations over equilibrium snapshots. In that case, another option is SampledCorrelations vs InstantSampledCorrelations, but it starts to feel long. I think a common use case will be to call InstantCorrelations on an energy minimizing configuration, so leaving out "sampled" makes sense to me.
Presumably InstantCorrelations can be looking at correlations beyond just dipole-dipole, so StructureFactor -> Correlations makes sense to me.
There was a problem hiding this comment.
What are the primary use cases of InstantCorrelations right now? Also note the overlap with print_wrapped_intensities.
|
Yep, the warnings are not your fault. It's some Documenter issue, has to do with the recently tweaked docs pipeline. I haven't found a way to resolve them, but I think they occur when you link from one docstring to another docstring. |
|
Gotcha. I think one warning will go away if a docstring is added to |
|
Per @ddahlbom's comment:
|
ddahlbom
force-pushed
the
SampledCorrelations
branch
from
e6021b2
to
3a0721f
Compare
docs/src/versions.md
Outdated
| `StructureFactor` type renamed `SampledCorrelations`. An appropriate | ||
| `SampledCorrelations` is created by calling either `dynamical_correlations` or | ||
| `instant_correlations` instead of `DynamicStructureFactor` or |
There was a problem hiding this comment.
Add hyperlinks,
[`SampledCorrelations`](@ref)
and similarly for dynamical_correlations and instant_correlations?
There was a problem hiding this comment.
Thanks. Added. (Note that the docstring for SampledCorrelations is basically just a stub and points to the docstrings for dynamical_correlations and instant_correlations. But I think it is useful to export SampledCorrelations, just because so many locations in the docs really want to refer to SampledCorrelations itself.)
|
Looks good to me. Thanks David! |
| @@ -46,7 +55,7 @@ can be resolved by passing an explicit `offset`. | |||
| The function [`remove_periodicity!`](@ref) disables periodicity along specified | |||
| dimensions. | |||
|
|
|||
| Rename `StaticStructureFactor` to [`InstantStructureFactor`](@ref). | |||
| Rename `StaticStructureFactor` to [`InstantStructureFactor`]. | |||
There was a problem hiding this comment.
This is gone now, right?
There was a problem hiding this comment.
Yes. That line is describing an earlier version, I thought. Just keeping the record while removing the broken hyperlink.
There was a problem hiding this comment.
(Guess I forgot to remove the square brackets.)
This just changes the name
StructureFactortoSampledCorrelations. It also eliminates the automatic generation of an initial sample when creating aSampledCorrelations, as was discussed some time back.I am getting some new warnings when building the docs, but they seem to look fine when I inspect them.
There's not much here, but many lines are touched. Let me know if you see any problems.