Chichi Lalescu authored 8 years ago and Cristian Lalescu committed 8 years ago

I want to properly accomodate postprocessing codes, so I need something
more general than `direct_numerical_simulation`.

doesn't do anything for now

Berenger Bramas authored 8 years ago and Cristian Lalescu committed 8 years ago

Add a sampling method -- update to be generic against the field -- need to update NSVEparticles::do_stats

for scaling tests, if we don't want to get stuck on I/O things...

keep NSVEparticles, since it's a child of NSVE and everything is more
consistent this way

Update particles output to limite the number of processes involved -- manage a minimum chunk size -- write a multiple of a chunk

also fix a typo in the vorticity equation wrapper, but the code fails
anyway, so that's not a big deal.
I still need to generate initial conditions.

I doubt that it does anything useful, but it compiles.

found
https://seasonofcode.com/posts/how-to-add-custom-build-steps-and-commands-to-setuppy.html
and used it to clean up the compiling step.

USE_TIMINGOUTPUT should be controlled from setup.py.
In this commit, I'm trying something elegant that's supposed to work,
but the "build" command actually fails for some reason.
Hence the commit, and I need to fix it in future commits.

the idea is to modify the fluid_solver class into something that uses
the field class, to the point where if we want to change the backend to
something other than plain FFTW, the new solver class will not need to
be modified in any way.

the "vorticity_equation" class included in this commit is not
functional, it is only a checkpoint.