Re: [h5md-user] Varying particle number

[Peter Colberg]

Hi Konrad,

On Tue, Oct 07, 2014 at 09:22:18AM +0200, Konrad Hinsen wrote:
> How would this number be interpreted in the most general case:
> (1) multiple subgroups in a trajectory, all of which could have
> a varying number of particles, (2) different kinds of particles,
> possible mixed in a subgroup, (3) partially overlapping subgroups?
> I actually suspect that some combination of 1-3 must be
> forbidden in order to make variable particle numbers possible.

particle_number would be a count of the particles in a subgroup, so
all three cases should be fine. The motivation is to know a priori
how many particles are stored in a subgroup; without requiring “id”
when particles are stored in a continguous fashion, where the empty
placeholders are at the end of each dataset.

Peter