Getting connected primitives in Python?

[simonj]

Hi everyone!

Yesterday I tried expr...

[DELETED]

Nevermind, sorry about the post, was dumb. Haha

regards/

Simon

edited away the question to not steal anyones time

Edited April 22, 2009 by simonj

[stevenong]

You can post your solution here if you got it.

Cheers!

steven

[simonj]

You can post your solution here if you got it.

Cheers!

steven

Yeah, sorry about that, hehe. I was going to but I had some quirks to work out.

Anyway, here is my solution, maybe there is a more effective one.

Since it doesn't seem like python (or more exactly "hython", right?) has a built in function, and I don't know how to use expressions inside of Python (any tips on that?) I created a Attribute Create-node

This node is set to create primitive attributes which will be a string. I checked "Write values" and put `primneighbours("../prim",$PR,2)` in the "string"-field. Note the ` before and after the expression, that makes sure that you don't put the expression as a string but the result of that expression instead.

This will set an attribute on each primitive which is a series of the connected primitives (aka the neighbours), for example, it can look like this for a primitive in the middle of the grid: 10 18 20 37.

Great, now all I have to do is to access these in python. Because the neighbours are in a string I need to fix that, python has a function called split that will do that for us.

so here is an example written in an IDLE.

>>> neighbours = "10 18 20 37"
>>> neighbours
'10 18 20 37'
>>> neighbours = neighbours.split()
>>> neighbours
['10', '18', '20', '37']

So you see that suddently we have these numbers in an array, and now they can be accessed through a loop:

in a python sop, this would be how you could do it:

geo = hou.pwd().geometry()

neighbours = geo.findPrimAttrib("neighbours")

for prim in geo.prims():
        nb = prim.attribValue(neighbours)
        for n in nb.split():
            num = int(n)
            nprim = geo.iterPrims()[num]

Let me explain it from the top.

geo = ... is a default to set in a sop node I think, makes access easier to write.

neighbours = ... is set to create a reference to an attribute instead of looking it up everytime in the loop, it's quicker this way.

for prim in geo.prims() iterates through all primitives

nb = ... gets the string containing the neighbours.

for n in nb.split(): put the strings numbers into an array and loops through them

num = int(n) converts the current number retrieved to an int instead of a string.

nprim = ... gets the primitive that has that id (geo.iterPrims()[what_id] is the prefered method for getting a specific primitive.)

So, that's it!

Any suggestions, or more efficient ways to do this are very welcome.

[edit, by the way, the reason I expained it so thoroughly is because it's a great way for me to remember it ]

Edited April 25, 2009 by simonj

[JoshJ]

just wondering, could you supply a hip or hipnc file with that setup? I tried to recreate it but I think I'm missing some crucial steps in the workflow. Thanks!

[simonj]

just wondering, could you supply a hip or hipnc file with that setup? I tried to recreate it but I think I'm missing some crucial steps in the workflow. Thanks!

Ofcourse I can, maybe I wasn't great at explaining either, a long explanation is not always a good one, hehe

I'm new at this .otl stuff, but I think this file should work "right out of the box".

This is a simple example and quite buggy, especially if you try lowering the "falloff" value below 2, you will see some wierd, and if you're lucky, cool, stuff.

A question for the pros though: How could I make this run "backwards" since now only the falloff value works "in front" of the primitives. and I would like it to work in all directions.

neighbours_python.zip