Skip to content

Constructors

In WGSL, matrices can be constructed or initialized using three primary constructor styles: zero-value initialization, column-vector composition, or element-by-element scalar initialization in column-major order.

Style Syntax Pattern Description
Zero-Value matCxRf()
matCxR<T>()		 Initializes all elements of the matrix to their default zero value (0.0).
Column-Wise matCxR<T>(c0, c1, ...) Constructs the matrix by supplying exactly C column vectors.
Scalar-Wise matCxR<T>(s0, s1, ...) Constructs the matrix from a flat sequence of scalars in column-major order.

Zero-Value Constructors

Constructs a matrix with all elements initialized to 0.0. You can use either the generic syntax or the convenient type aliases (like mat3x4f).

let m_zero  = mat3x4f();       // Equivalent to mat3x4<f32>() with all 0.0s
let m_f32   = mat2x2<f32>();   // 2x2 matrix filled with 0.0

Column-Wise Constructors

Constructs a matrix by supplying exactly C column vectors, where each column vector has exactly R components. This is the most common way to compose matrices.

Because WGSL is column-major, the first vector argument becomes the first (leftmost) column of the matrix, the second argument becomes the second column, and so on.

let col_0 = vec2f(1.0, 2.0);
let col_1 = vec2f(3.0, 4.0);
let col_2 = vec2f(5.0, 6.0);

// Composes a 3x2 matrix from three column vectors
let m_cols = mat3x2f(col_0, col_1, col_2);

Scalar-Wise Constructors

Constructs a matrix from a flat list of individual scalar values. To specify the scalar values, they must be supplied in column-major order: the elements of the first column are specified first (top-to-bottom), followed by the elements of the second column, and so forth.

For a matrix with C columns and R rows, you must supply exactly \(C \times R\) scalar arguments.

For example, mat2x3f(a, b, c, d, e, f) constructs a 2-column, 3-row matrix of f32s, where the first column is populated by (a, b, c) and the second column is populated by (d, e, f):

// Constructing a 2x3 matrix (2 columns, 3 rows) using 6 scalars
let m_scalars = mat2x3f(1.0, 2.0, 3.0, 4.0, 5.0, 6.0);

The resulting matrix layout is:

╭           ╮
│ 1.0   4.0 │
│ 2.0   5.0 │
│ 3.0   6.0 │
╰           ╯