Per particle position

[pinxue]

Is it possible to query the position of an individual particle in a particle object. For example, How would I find the position for particle1.pt[4]....I need to put all the individual particle positions into an array. Thanks in advance....

[joojaa]

Yes you can do it with getattr but its slow and shoulnt be used unless absolutely neccesery. its much easier to use global arrays to store tha data on the fly and use that in calculations.

[pinxue]

Well, the ammount of particles is very small per particle object, Under 30 to 40. I don't think that would cause too much of a slow down, or would it?

[pinxue]

I've been playin around with the getAttr command and still can't get it to return the position. Could you maybe explain how to do it or tell me a little more about the global array way...?

[robin-scher]

I'm not sure I understand exactly what you are trying to do.

You can get the position of an individual particle in the run-time expression of the particle shape simply by accessing the position attribute. if you created a bunch of particles, and then wanted to store their positions in an array, you could write a run-time expression like this:

global vector $pos[];
$pos[particleId] = position;

then step forward one frame, and the runtime expression would fill in the array automatically indexed by particle ID. This may be a really bad idea if you have a lot of particles.

If you are trying to do something like generate geometry at the particle positions, or move existing geometry to the particle position, you could use the runtime expression and reference the particleShape.position attribute directly, rather than trying to store it.

The best solution depends on what, exactly, you want to do with the particle positions.

[pinxue]

Well I'm still not entirely sure what I am trying to do, but so far it goes like this. I have a scene and I want to place a lot of rocks around the bases of certain cliffs. What I would like to do is use the particle tool, and just put a particle where I want a rock to eventually go. So I place say 25 particles randomly around the base of this cliff. I then want to be able to model 10-15 different rocks, varying in shape and scale. I would then like to instance those rocks to random particles around the base of the cliff. note the particles never move. But here's where I think it will get a little difficult. I want this script to look at the particle positions in relation to one another. And the ninstance the various sized rocks accordingly. For Example, if there are a couple particles bunched together, I only want small rocks there so they are not likely to penetrate one another and if there is a good distance between particles, Then I want a larger rock placed there. I know it would be alot simpler just to model the rocks and duplicate/translate them into the proper positions, but I have got to learn mel somehow and I would really like to make a script with possibly a window interface to accomplish this task(Of course the window interface will come after the script actually works) So, that is basically what I'm trying to do, Hope it doesn't sound to complex for a simple task, but I would imagine once it's programed, it would save me alot of time. I have a lot of cliffs with tons of rocks to be placed. Thanks to anyone that can help me in this situation....

[tbriggs]

While you can use the position command within a particle expression to get the values, from outside an expresson use the mel command:
particle -attribute position -order $n -query particle1 ,
where order is the number of the particle's array index. You could use count or particleId instead of order if you want. As to the idea of testing the distance between particles and scaling instances accordingly, you can, but bear in mind that the number of tests rises with the square of the particle count, and would quickly become as they say, computationally intractable. This might be best done by eye.

[pinxue]

Thanks very much for the help. I know it would have been way easier to do this by eye, I have already spent two days trying to figure out ways to write the script. I would have probably been done setting up the whole scene by now!...Anyways, I'll probably put this idea on the back burner for now and just continue on with my scene.