I thought group pops were so simple. Just define some groups with rules or by their generator and voila. But this has not been the case. My groups don't behave! Members are changing out of groups like crazy and it will happen at some unpredictable frame that I can't figure out. The preserve group checkbox seems clear enough but the problem comes with groups that have intersecting members. Obviously, I'm just assigning members improperly. Can anyone clarify for me the use of the Combine tab in the group pop?
I've included images of my group pop create and combine tabs. Thanks!

Don't read this unless you need more detail and enjoy tedium! I'm attempting to create a monster out of a vortex of falling snow. Here are some details behind my groups.
I've defined many particle groups, but the most important are the 4 following:

near_monster: a bounding box around the monster where I eventually want a vortex.

far_from monster: using the combine tab in the group node for near_monster simply to make a group of points not-equal to near_monster. So I believe this just means everything outside the bounding box. I want these particles to drift normally like falling snow so they aren't affected by the vortex since they are farther from it.

self: a group of particles generated by the monster itself. Since the monster is inside the bounding box, it is being included in near_monster. I want to subtract these from near_monster.

spline: a group of particles generated on a curve that serve as the center particles for an orbit node. This is how I'm creating the vortex. I want the near_monster to spin around these particles. This is also in the bounding box for near_monster so they need to be subtracted from near_monster also.

this may not be as elegant as i could be, but what about assigning different groups colors and then selecting them based on color.
you could add and subtract values from each other.
maybe near_monster is CR = 1 then subtract .2 for the self.

hope this helps

-jacob

Thanks for the interesting option. I'll give that a try. However, despite my messy question (I could have stated it much more concisely I see now), I have no manual with Houdini (NC) and was hoping to learn more about the appropriate use of the Group Pop. It seems so clear about the use of subtracting a group, yet it doesn't behave as I anticipated. I'm not even sure which group they are being subtracted from. Also, to subtract a group why do you have to create a =! group or = group under the combine tab in the first place. Oh well, its a mystery. Thanks, again.

bradwallace
I thought group pops were so simple. Just define some groups with rules or by their generator and voila. But this has not been the case. My groups don't behave! Members are changing out of groups like crazy and it will happen at some unpredictable frame that I can't figure out. The preserve group checkbox seems clear enough but the problem comes with groups that have intersecting members. Obviously, I'm just assigning members improperly. Can anyone clarify for me the use of the Combine tab in the group pop?
I've included images of my group pop create and combine tabs. Thanks!

When you use a bounding box as a grouping selector, you are subject to particles entering or leaving the bounding box, and ergo, your inverse group will follow right along. Furthermore, if your particles are flying in circles, they could easily pop out of the bbox and then back in a frame or two later, as the circle goes inside the bbox again.

Now I'm surprised that preserve-groups isn't working, as it should alter the group POP to always add the particles to the group. I would try a simple experiment to see if the button behaves as you think it should, e.g. make a couple of particles travel in a horizontal line, and see if the bounding box captures them, and if they stay in.

The combine section lets you use booleans to create new groups. Your setup seems ok: start with the inverse group, and then subtract a couple of others. Do you maybe not have preseve-groups clicked on on one of those? That would cause your group flicker.

– Antoine

Thanks, Antoine for helping clarify. I'll do some more testing. I still find the combine tab a little obscure. It seems you must define a second inverse or equal group in order to use the booleans. Strange.