python-vector-swizzling

The vector_swizzling library builds on top of numpy to provide flexible and intuitive vector manipulation with swizzling capabilities, designed to resemble GLSL vector handling for 2D, 3D, and 4D vectors.

Overview

SVec is a base class for representing vectors with a numpy array and a lookup table allowing swizzling. Swizzling lets you access components in different combinations or orders, similar to GLSL's swizzling. See Swizzling on Wikipedia for more information.

Classes and Structure

The main SVec class has three subclasses for specific dimensions:

• SVec2: for 2D vectors.
• SVec3: for 3D vectors.
• SVec4: for 4D vectors.

These subclasses come with dimension-specific operations, such as srotate for SVec2 and scross for SVec3.

Component Lookup Table

The swizzling table maps to the following indices:

• 'x' or 'r' = 0
• 'y' or 'g' = 1
• 'z' or 'b' = 2
• 'w' or 'a' = 3

This table lets you use letters from both Cartesian (x, y, z, w) and color (r, g, b, a) spaces interchangeably.

Usage Examples

Declaring Vectors

You can declare vectors in different ways:

# Standard declaration
a = SVec2(1, 2)
b = SVec3(1, 2, 3)
c = SVec4(1, 2, 3, 4)
d = SVec4(4, 3, 2, 1)

# Using list as components
a = SVec2([1, 2])
b = SVec3([1, 2], 3)
c = SVec4(1, 2, [3, 4])
d = SVec4([4, 3, 2], 1)

# Using vectors as components
b = SVec3(a, 3)
c = SVec4(b, 4)`
d = SVec4(c.wzy, 1)

Swizzling Vectors

You can assign to and operate on swizzled vectors, as well passing them as function arguments:

# Assigning and operating on swizzled vectors
b.xy = a.xy + b.yx

# Swizzling with RGBA space
d.xyzw = d.rgba

# Swizzling an SVec2 into an SVec3 and calling scross
cross_vec = scross(a.xyx, b)

Swizzling quirks

This module allows higher dimensional swizzles and swizzle chaining:

# Swizzle an SVec4 into a 7D SVec and call a dimension agnostic function on it
normalized_7D_vector = snormalize(d.xyzwxyz)

# Swizzle chaining is permitted but only works for the first 4 components, just like GLSL
d.xyzwxyz.xyzw
            #^ You run out of swizzles here

Vectors as lists

Vectors are just cool lists, so you can use them as such

# Iterate over vector components
for i in d.xyzwxyz:
    print(i)

# Assign to a component with its index
d[3] = 1