I don’t know what you mean by problems or error.
The binning logic does not account for periodic BC.
So you won’t get an answer that includes that.
Are you saying there is some other problem with the current code, or
simply that you don’t get the answer that includes PBC?
It would be possible to add a flag, like pbc yes/no, to
the compute chunk/atom command that accounted
for PBC when using bin/sphere or bin/cylinder. It would
probably also make sense in those cases to enforce
that the specified radmax < 1/2 box size. Is that
the use case you are wanting?
You also mention having the sphere center be dynamic,
e.g. move with an atom. There is also no logic currently
to do that, though we’ve thought about enabling it
by use of equal-style variables for specifying the bin
geometry.
Thank you for clarifying, I see what you mean now. Yes, I just meant I do not get the answer including PBC, which is what I was expecting at first, but is consistent with the binning logic.
The use case I have in mind is one fixed (or actually rather moving, but it should be the same in terms of binning) heat-source-particle (or several of them) around which I want to measure the temperature profile inside a box with periodic boundary conditions. That would then actually be done best by using the pbc yes/no flag you suggested to not interfere with the usual binning logic.
I modified bin/sphere for my specific use case to read in a group consisting of just one atom and updating "sorigin" with its current position whenever binning is done, as well as applying minimum image convention. Using equal-style variables and a pbc yes/no flag as you mentioned sounds cleaner though.