[REQUEST] C and Python Modulus operator for vectorial quantities

[fritzio]

Description

• The modulus operator mod is currently only defined for scalar, i.e., float quantities. I suggest implementing it for vectorial and tensorial quantities, e.g., vec3, so that we can omit a wrapper function, such as,

@wp.func
def mod(a: wp.vec3, b: wp.vec3) -> wp.vec3:
    return wp.vec3(wp.mod(a[0], b[0]), wp.mod(a[1], b[1]), wp.mod(a[2], b[2]))

• The modulus operator mod gives the modulus as defined in C and not in Python, which might lead to surprising behavior if we are unaware. I suggest adding a Pythonic modulus operator, e.g.,

@wp.func
def pymod(a: float, b: float) -> float:
    return wp.mod(wp.mod(a, b) + b, b)

Context

The motivation for this request is to have less boilerplate code in our application.

[sudo-panda]

Can I take this up?

[shi-eric]

I think @christophercrouzet is actively working on this issue already.

[christophercrouzet]

We just merged the changes to the branch main. So far we only added support for vectors to wp.mod() and documented the built-in as being implemented as a truncated division to alleviate the ambiguity with Python's implementation. We haven't implemented the floored division variant for now.