Re: [h5md-user] Particle tracking

[Peter Colberg]

On Mon, Sep 02, 2013 at 01:31:07PM -0400, Peter Colberg wrote:
> There seems to be no way to get around an O(N) operation, since the
> elements of the id dataset have to be iterated due to the possible
> presence of holes along the dimension of the particle number (see the
> remarks on MPI-parallel simulations).
> 
> A trajectory subgroup dataset containing the instantaneous particle
> counts would help with memory allocation though, when buffers are to
> be allocated according to the instantaneous particle number, rather
> than the maximum particle number.

Then again, for parallel reading, a global particle count would not be
useful, whereas the (implementation-dependent) particle count per
processor would be.

The “counting ids” solution may not be the most efficient, but it is
generally applicable, until the H5MD community gains experience with
large-scale datasets (e.g., »10⁶ particles) and parallel writing and
reading.

Peter