Particle masses as *gasp* masses

[Zephyr Penoyre]

Hi all,

  Greg and I have been talking about particle masses (particularly active particles in enzo 3+) and how now might be the perfect time to drop the strange convention of masses being pseudo-densities and instead record actual particle mass.

  There are numerous advantages to this, especially for active particles where the mass is much more often used than the density contribution (and where density is needed is entirely confined to grid routines where it is easily calculated) and this means we no longer need to worry about local properties for particle masses.

  For non-active particles the case is less clear cut. I'd be intrigued to hear if people think the convenience and clarity of masses being masses outweighs the slight extra calculation in the gravity solvers?

  It seems to be a sensible transition to make, and a good time to do it (going forward with it only in enzo 3 and leaving the 2 untouched), but we were keen to gauge people's reactions, fears, hopes and dreams. Obviously it may also require small changes to people's own codes if implemented across enzo 3, but I don't have a good sense of how intractable a problem that is?

     Keen to hear any thoughts and potential pitfalls anyone can think of,

       Best,

         Zephyr

[Matthew Turk]

Hi Zephyr,

In general, I am +100 on making this change for active particles.  I'm +0 on changing non-active particles.  If we/you did this, I'd like to lobby that we somehow distinguish the existing datasets from new ones; usually we do this by using a different name for the data field...  unfortunately, we already call it particle mass, if I remember correctly.

If we could just get a new field name so that we didn't accidentally look at the masses and wonder what we're looking at, that'd assuage my concerns.  Otherwise, I think this is great since it's the much more natural way of storing and dealing with data.

One possible way to make this transition more seamless is to implement UpdateDensity() and UpdateMass() methods, switch everything that modifies this to one or the other, then switch those places.

-Matt

--
You received this message because you are subscribed to the Google Groups "enzo-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to enzo-dev+u...@googlegroups.com.
To post to this group, send email to enzo...@googlegroups.com.
Visit this group at https://groups.google.com/group/enzo-dev.
For more options, visit https://groups.google.com/d/optout.

[Britton Smith]

Hi Zephyr,

I'm also totally on board with this.  I agree with Matt that this it would be extremely nice if there were some distinguishing feature in the datasets so that on the analysis side we aren't left guessing.  If there is such a way, then I would be in favor of converting both active and non-active particles, just for consistency.

Britton

[Brian O'Shea]

Hi folks,

I'm following up on this thread.  Zephyr issued a pull request a while ago to convert article particle non-masses (i.e., densities) to actual masses, and has been patiently waiting for people to look at it.  The PR can be found here:

    https://bitbucket.org/enzo/enzo-3.0/pull-requests/83/changing-active-particle-masses-to-actual/diff

I've asked him to come up with some sort of test that verifies the modified code is behaving correctly (see comments in the PR), and I am also going to look through the source code.  It would be very helpful if a couple of people can agree to look at this PR and comment on it.

Thank you!

Brian