Struct Transform 

pub struct Transform { /* private fields */ }

A two-dimensional affine transform in double-precision design coordinates.

Coefficients are column-major [xx, yx, xy, yy, translation_x, translation_y]. Positive rotation is clockwise in the downward-positive coordinate system. Self::then applies transforms in reading order, not matrix-product order. Device-space geometry retains these f64 coefficients until explicit storage boundaries.

Implementations

impl Transform

pub const IDENTITY: Self

Leaves positions and vectors unchanged

pub const fn from_coefficients(coefficients: [f64; 6]) -> Self

Preserves raw affine coefficients; consuming geometry validates them.

pub const fn coefficients(self) -> [f64; 6]

Returns column-major affine coefficients without reducing precision.

pub const fn translate(horizontal: f64, vertical: f64) -> Self

Translates positions without changing vectors.

pub fn rotate(radians: f64) -> Self

Rotates clockwise around the coordinate origin, in radians.

pub const fn scale(factor: f64) -> Self

Uniformly scales around the coordinate origin; negative values reflect.

pub const fn scale_non_uniform(horizontal: f64, vertical: f64) -> Self

Scales each axis independently around the coordinate origin.

pub fn skew(horizontal: f64, vertical: f64) -> Self

Skews each axis by the tangent of its angle in radians. Both shears act in one matrix, rather than as sequential operations.

pub fn then(self, next: Self) -> Self

Applies this transform first, followed by next.

use motion_core::{Point, Transform};
let transform = Transform::translate(10.0, 0.0).then(Transform::scale(2.0));
assert_eq!(transform.apply(Point::new(1.0, 3.0)), Point::new(22.0, 6.0));

pub fn apply(self, point: Point) -> Point

Maps a position, including translation.

pub fn apply_vector(self, vector: Vector2) -> Vector2

Maps a displacement, ignoring translation.

pub fn is_finite(self) -> bool

Returns whether all affine coefficients are finite.

pub fn inverse(self) -> Option<Self>

Inverts a finite nonsingular transform when its inverse remains finite. Returns None for singular/nonfinite inputs or unrepresentable results. Rows are scaled before computing the determinant to avoid ordinary underflow/overflow for extremely small, large or unequal axis scales.

pub fn decompose(self) -> Option<TransformComponents>

Decomposes into scale, horizontal shear, rotation, then translation.

The horizontal scale is positive; reflection is carried by vertical scale. Returns None for singular/nonfinite inputs or nonfinite components. Shear is stored as a coefficient, not an angle, to avoid losing precision when a large tangent is rounded to a right angle.

pub fn interpolate(&self, target: &Self, progress: f64) -> Self

Interpolates decomposed components and takes the shortest rotation arc. A half-turn tie rotates clockwise; progress may extrapolate. Exact zero/one return the original matrices. If either endpoint cannot be decomposed, coefficients interpolate directly, allowing singular endpoints such as zero-scale entrances. Reflection changes can pass through a singular scale.

Trait Implementations

impl Clone for Transform

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for Transform

impl Debug for Transform

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Transform

fn default() -> Self

Returns the “default value” for a type.

impl Interpolate for Transform

fn interpolate(&self, target: &Self, progress: f64) -> Self

Blends toward target with nominal progress zero at self and one at target.

impl PartialEq for Transform

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Transform