fix heat group has no atoms?

please *always* copy the mailing list on your replies.
i only answer to mails sent to me personally, if they
concern me personally, or code that i wrote. this is
not such a case. thank you.

Hi, Dr.Axel Kohlmeyer,
   I am still confusing. But the syntax of fix heat command: fix ID group-ID heat N eflux region region-ID . So, i need to use group ID into this fix. Since groups are static, then How to define unstatic groups? After all, the atoms are moving from groups to groups.

no, you don't need to define "changing group".
think about it: you are not the first person to
set up this kind of calculation. if there was such
a tricky problem, the documentation would
certainly mention it, would it not?

atoms a *not* moving from group to group,
they are moving from region to region.

   Because i only want add energy to hot region and subtract energy from cold1 and cold2 region. I tried "fix heat N eflux region region-ID " without group-ID. It still couldnt pass through in Lammps.

of course not. if you want to restrict atoms only
by region, you use the predefined group all.
as the manual says, you *have* to provide a group.
if you don't want/need to restrict atoms in a static
fashion (e.g. by atom type; for which you *can* use
static definition) then you just use "all"

LAMMPS is a compute program and doesn't have
any common sense, it just follows the rules and
you have to keep that in mind when formulating
input files.