nalgebra/src/geometry/rotation.rs
Benjamin Saunders 0541f13b26 Concise Debug impls
Replace the verbose derived (or nearly equivalent) Debug impls for
several newtypes with explicit impls that forward to the inner type,
making readable diagnostics logging much easier.
2021-09-12 10:56:25 -07:00

606 lines
21 KiB
Rust
Executable File

use approx::{AbsDiffEq, RelativeEq, UlpsEq};
use num::{One, Zero};
use std::fmt;
use std::hash;
#[cfg(feature = "abomonation-serialize")]
use std::io::{Result as IOResult, Write};
#[cfg(feature = "serde-serialize-no-std")]
use serde::{Deserialize, Deserializer, Serialize, Serializer};
#[cfg(feature = "serde-serialize-no-std")]
use crate::base::storage::Owned;
#[cfg(feature = "abomonation-serialize")]
use abomonation::Abomonation;
use simba::scalar::RealField;
use simba::simd::SimdRealField;
use crate::base::allocator::Allocator;
use crate::base::dimension::{DimNameAdd, DimNameSum, U1};
use crate::base::{Const, DefaultAllocator, OMatrix, SMatrix, SVector, Scalar, Unit};
use crate::geometry::Point;
/// A rotation matrix.
///
/// This is also known as an element of a Special Orthogonal (SO) group.
/// The `Rotation` type can either represent a 2D or 3D rotation, represented as a matrix.
/// For a rotation based on quaternions, see [`UnitQuaternion`](crate::UnitQuaternion) instead.
///
/// Note that instead of using the [`Rotation`](crate::Rotation) type in your code directly, you should use one
/// of its aliases: [`Rotation2`](crate::Rotation2), or [`Rotation3`](crate::Rotation3). Though
/// keep in mind that all the documentation of all the methods of these aliases will also appears on
/// this page.
///
/// # Construction
/// * [Identity <span style="float:right;">`identity`</span>](#identity)
/// * [From a 2D rotation angle <span style="float:right;">`new`…</span>](#construction-from-a-2d-rotation-angle)
/// * [From an existing 2D matrix or rotations <span style="float:right;">`from_matrix`, `rotation_between`, `powf`…</span>](#construction-from-an-existing-2d-matrix-or-rotations)
/// * [From a 3D axis and/or angles <span style="float:right;">`new`, `from_euler_angles`, `from_axis_angle`…</span>](#construction-from-a-3d-axis-andor-angles)
/// * [From a 3D eye position and target point <span style="float:right;">`look_at`, `look_at_lh`, `rotation_between`…</span>](#construction-from-a-3d-eye-position-and-target-point)
/// * [From an existing 3D matrix or rotations <span style="float:right;">`from_matrix`, `rotation_between`, `powf`…</span>](#construction-from-an-existing-3d-matrix-or-rotations)
///
/// # Transformation and composition
/// Note that transforming vectors and points can be done by multiplication, e.g., `rotation * point`.
/// Composing an rotation with another transformation can also be done by multiplication or division.
/// * [3D axis and angle extraction <span style="float:right;">`angle`, `euler_angles`, `scaled_axis`, `angle_to`…</span>](#3d-axis-and-angle-extraction)
/// * [2D angle extraction <span style="float:right;">`angle`, `angle_to`…</span>](#2d-angle-extraction)
/// * [Transformation of a vector or a point <span style="float:right;">`transform_vector`, `inverse_transform_point`…</span>](#transformation-of-a-vector-or-a-point)
/// * [Transposition and inversion <span style="float:right;">`transpose`, `inverse`…</span>](#transposition-and-inversion)
/// * [Interpolation <span style="float:right;">`slerp`…</span>](#interpolation)
///
/// # Conversion
/// * [Conversion to a matrix <span style="float:right;">`matrix`, `to_homogeneous`…</span>](#conversion-to-a-matrix)
///
#[repr(C)]
pub struct Rotation<T, const D: usize> {
matrix: SMatrix<T, D, D>,
}
impl<T: fmt::Debug, const D: usize> fmt::Debug for Rotation<T, D> {
fn fmt(&self, formatter: &mut fmt::Formatter<'_>) -> Result<(), fmt::Error> {
self.matrix.fmt(formatter)
}
}
impl<T: Scalar + hash::Hash, const D: usize> hash::Hash for Rotation<T, D>
where
<DefaultAllocator as Allocator<T, Const<D>, Const<D>>>::Buffer: hash::Hash,
{
fn hash<H: hash::Hasher>(&self, state: &mut H) {
self.matrix.hash(state)
}
}
impl<T: Scalar + Copy, const D: usize> Copy for Rotation<T, D> where
<DefaultAllocator as Allocator<T, Const<D>, Const<D>>>::Buffer: Copy
{
}
impl<T: Scalar, const D: usize> Clone for Rotation<T, D>
where
<DefaultAllocator as Allocator<T, Const<D>, Const<D>>>::Buffer: Clone,
{
#[inline]
fn clone(&self) -> Self {
Self::from_matrix_unchecked(self.matrix.clone())
}
}
#[cfg(feature = "bytemuck")]
unsafe impl<T, const D: usize> bytemuck::Zeroable for Rotation<T, D>
where
T: Scalar + bytemuck::Zeroable,
SMatrix<T, D, D>: bytemuck::Zeroable,
{
}
#[cfg(feature = "bytemuck")]
unsafe impl<T, const D: usize> bytemuck::Pod for Rotation<T, D>
where
T: Scalar + bytemuck::Pod,
SMatrix<T, D, D>: bytemuck::Pod,
{
}
#[cfg(feature = "abomonation-serialize")]
impl<T, const D: usize> Abomonation for Rotation<T, D>
where
T: Scalar,
SMatrix<T, D, D>: Abomonation,
{
unsafe fn entomb<W: Write>(&self, writer: &mut W) -> IOResult<()> {
self.matrix.entomb(writer)
}
fn extent(&self) -> usize {
self.matrix.extent()
}
unsafe fn exhume<'a, 'b>(&'a mut self, bytes: &'b mut [u8]) -> Option<&'b mut [u8]> {
self.matrix.exhume(bytes)
}
}
#[cfg(feature = "serde-serialize-no-std")]
impl<T: Scalar, const D: usize> Serialize for Rotation<T, D>
where
Owned<T, Const<D>, Const<D>>: Serialize,
{
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where
S: Serializer,
{
self.matrix.serialize(serializer)
}
}
#[cfg(feature = "serde-serialize-no-std")]
impl<'a, T: Scalar, const D: usize> Deserialize<'a> for Rotation<T, D>
where
Owned<T, Const<D>, Const<D>>: Deserialize<'a>,
{
fn deserialize<Des>(deserializer: Des) -> Result<Self, Des::Error>
where
Des: Deserializer<'a>,
{
let matrix = SMatrix::<T, D, D>::deserialize(deserializer)?;
Ok(Self::from_matrix_unchecked(matrix))
}
}
impl<T, const D: usize> Rotation<T, D> {
/// Creates a new rotation from the given square matrix.
///
/// The matrix orthonormality is not checked.
///
/// # Example
/// ```
/// # use nalgebra::{Rotation2, Rotation3, Matrix2, Matrix3};
/// # use std::f32;
/// let mat = Matrix3::new(0.8660254, -0.5, 0.0,
/// 0.5, 0.8660254, 0.0,
/// 0.0, 0.0, 1.0);
/// let rot = Rotation3::from_matrix_unchecked(mat);
///
/// assert_eq!(*rot.matrix(), mat);
///
///
/// let mat = Matrix2::new(0.8660254, -0.5,
/// 0.5, 0.8660254);
/// let rot = Rotation2::from_matrix_unchecked(mat);
///
/// assert_eq!(*rot.matrix(), mat);
/// ```
#[inline]
pub const fn from_matrix_unchecked(matrix: SMatrix<T, D, D>) -> Self {
Self { matrix }
}
}
/// # Conversion to a matrix
impl<T: Scalar, const D: usize> Rotation<T, D> {
/// A reference to the underlying matrix representation of this rotation.
///
/// # Example
/// ```
/// # use nalgebra::{Rotation2, Rotation3, Vector3, Matrix2, Matrix3};
/// # use std::f32;
/// let rot = Rotation3::from_axis_angle(&Vector3::z_axis(), f32::consts::FRAC_PI_6);
/// let expected = Matrix3::new(0.8660254, -0.5, 0.0,
/// 0.5, 0.8660254, 0.0,
/// 0.0, 0.0, 1.0);
/// assert_eq!(*rot.matrix(), expected);
///
///
/// let rot = Rotation2::new(f32::consts::FRAC_PI_6);
/// let expected = Matrix2::new(0.8660254, -0.5,
/// 0.5, 0.8660254);
/// assert_eq!(*rot.matrix(), expected);
/// ```
#[inline]
#[must_use]
pub fn matrix(&self) -> &SMatrix<T, D, D> {
&self.matrix
}
/// A mutable reference to the underlying matrix representation of this rotation.
#[inline]
#[deprecated(note = "Use `.matrix_mut_unchecked()` instead.")]
pub unsafe fn matrix_mut(&mut self) -> &mut SMatrix<T, D, D> {
&mut self.matrix
}
/// A mutable reference to the underlying matrix representation of this rotation.
///
/// This is suffixed by "_unchecked" because this allows the user to replace the
/// matrix by another one that is non-inversible or non-orthonormal. If one of
/// those properties is broken, subsequent method calls may return bogus results.
#[inline]
pub fn matrix_mut_unchecked(&mut self) -> &mut SMatrix<T, D, D> {
&mut self.matrix
}
/// Unwraps the underlying matrix.
///
/// # Example
/// ```
/// # use nalgebra::{Rotation2, Rotation3, Vector3, Matrix2, Matrix3};
/// # use std::f32;
/// let rot = Rotation3::from_axis_angle(&Vector3::z_axis(), f32::consts::FRAC_PI_6);
/// let mat = rot.into_inner();
/// let expected = Matrix3::new(0.8660254, -0.5, 0.0,
/// 0.5, 0.8660254, 0.0,
/// 0.0, 0.0, 1.0);
/// assert_eq!(mat, expected);
///
///
/// let rot = Rotation2::new(f32::consts::FRAC_PI_6);
/// let mat = rot.into_inner();
/// let expected = Matrix2::new(0.8660254, -0.5,
/// 0.5, 0.8660254);
/// assert_eq!(mat, expected);
/// ```
#[inline]
pub fn into_inner(self) -> SMatrix<T, D, D> {
self.matrix
}
/// Unwraps the underlying matrix.
/// Deprecated: Use [`Rotation::into_inner`] instead.
#[deprecated(note = "use `.into_inner()` instead")]
#[inline]
pub fn unwrap(self) -> SMatrix<T, D, D> {
self.matrix
}
/// Converts this rotation into its equivalent homogeneous transformation matrix.
///
/// This is the same as `self.into()`.
///
/// # Example
/// ```
/// # use nalgebra::{Rotation2, Rotation3, Vector3, Matrix3, Matrix4};
/// # use std::f32;
/// let rot = Rotation3::from_axis_angle(&Vector3::z_axis(), f32::consts::FRAC_PI_6);
/// let expected = Matrix4::new(0.8660254, -0.5, 0.0, 0.0,
/// 0.5, 0.8660254, 0.0, 0.0,
/// 0.0, 0.0, 1.0, 0.0,
/// 0.0, 0.0, 0.0, 1.0);
/// assert_eq!(rot.to_homogeneous(), expected);
///
///
/// let rot = Rotation2::new(f32::consts::FRAC_PI_6);
/// let expected = Matrix3::new(0.8660254, -0.5, 0.0,
/// 0.5, 0.8660254, 0.0,
/// 0.0, 0.0, 1.0);
/// assert_eq!(rot.to_homogeneous(), expected);
/// ```
#[inline]
#[must_use]
pub fn to_homogeneous(&self) -> OMatrix<T, DimNameSum<Const<D>, U1>, DimNameSum<Const<D>, U1>>
where
T: Zero + One,
Const<D>: DimNameAdd<U1>,
DefaultAllocator: Allocator<T, DimNameSum<Const<D>, U1>, DimNameSum<Const<D>, U1>>,
{
// We could use `SMatrix::to_homogeneous()` here, but that would imply
// adding the additional traits `DimAdd` and `IsNotStaticOne`. Maybe
// these things will get nicer once specialization lands in Rust.
let mut res = OMatrix::<T, DimNameSum<Const<D>, U1>, DimNameSum<Const<D>, U1>>::identity();
res.fixed_slice_mut::<D, D>(0, 0).copy_from(&self.matrix);
res
}
}
/// # Transposition and inversion
impl<T: Scalar, const D: usize> Rotation<T, D> {
/// Transposes `self`.
///
/// Same as `.inverse()` because the inverse of a rotation matrix is its transform.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use nalgebra::{Rotation2, Rotation3, Vector3};
/// let rot = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// let tr_rot = rot.transpose();
/// assert_relative_eq!(rot * tr_rot, Rotation3::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(tr_rot * rot, Rotation3::identity(), epsilon = 1.0e-6);
///
/// let rot = Rotation2::new(1.2);
/// let tr_rot = rot.transpose();
/// assert_relative_eq!(rot * tr_rot, Rotation2::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(tr_rot * rot, Rotation2::identity(), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use = "Did you mean to use transpose_mut()?"]
pub fn transpose(&self) -> Self {
Self::from_matrix_unchecked(self.matrix.transpose())
}
/// Inverts `self`.
///
/// Same as `.transpose()` because the inverse of a rotation matrix is its transform.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use nalgebra::{Rotation2, Rotation3, Vector3};
/// let rot = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// let inv = rot.inverse();
/// assert_relative_eq!(rot * inv, Rotation3::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(inv * rot, Rotation3::identity(), epsilon = 1.0e-6);
///
/// let rot = Rotation2::new(1.2);
/// let inv = rot.inverse();
/// assert_relative_eq!(rot * inv, Rotation2::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(inv * rot, Rotation2::identity(), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use = "Did you mean to use inverse_mut()?"]
pub fn inverse(&self) -> Self {
self.transpose()
}
/// Transposes `self` in-place.
///
/// Same as `.inverse_mut()` because the inverse of a rotation matrix is its transform.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use nalgebra::{Rotation2, Rotation3, Vector3};
/// let rot = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// let mut tr_rot = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// tr_rot.transpose_mut();
///
/// assert_relative_eq!(rot * tr_rot, Rotation3::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(tr_rot * rot, Rotation3::identity(), epsilon = 1.0e-6);
///
/// let rot = Rotation2::new(1.2);
/// let mut tr_rot = Rotation2::new(1.2);
/// tr_rot.transpose_mut();
///
/// assert_relative_eq!(rot * tr_rot, Rotation2::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(tr_rot * rot, Rotation2::identity(), epsilon = 1.0e-6);
/// ```
#[inline]
pub fn transpose_mut(&mut self) {
self.matrix.transpose_mut()
}
/// Inverts `self` in-place.
///
/// Same as `.transpose_mut()` because the inverse of a rotation matrix is its transform.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use nalgebra::{Rotation2, Rotation3, Vector3};
/// let rot = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// let mut inv = Rotation3::new(Vector3::new(1.0, 2.0, 3.0));
/// inv.inverse_mut();
///
/// assert_relative_eq!(rot * inv, Rotation3::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(inv * rot, Rotation3::identity(), epsilon = 1.0e-6);
///
/// let rot = Rotation2::new(1.2);
/// let mut inv = Rotation2::new(1.2);
/// inv.inverse_mut();
///
/// assert_relative_eq!(rot * inv, Rotation2::identity(), epsilon = 1.0e-6);
/// assert_relative_eq!(inv * rot, Rotation2::identity(), epsilon = 1.0e-6);
/// ```
#[inline]
pub fn inverse_mut(&mut self) {
self.transpose_mut()
}
}
/// # Transformation of a vector or a point
impl<T: SimdRealField, const D: usize> Rotation<T, D>
where
T::Element: SimdRealField,
{
/// Rotate the given point.
///
/// This is the same as the multiplication `self * pt`.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use std::f32;
/// # use nalgebra::{Point3, Rotation2, Rotation3, UnitQuaternion, Vector3};
/// let rot = Rotation3::new(Vector3::y() * f32::consts::FRAC_PI_2);
/// let transformed_point = rot.transform_point(&Point3::new(1.0, 2.0, 3.0));
///
/// assert_relative_eq!(transformed_point, Point3::new(3.0, 2.0, -1.0), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use]
pub fn transform_point(&self, pt: &Point<T, D>) -> Point<T, D> {
self * pt
}
/// Rotate the given vector.
///
/// This is the same as the multiplication `self * v`.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use std::f32;
/// # use nalgebra::{Rotation2, Rotation3, UnitQuaternion, Vector3};
/// let rot = Rotation3::new(Vector3::y() * f32::consts::FRAC_PI_2);
/// let transformed_vector = rot.transform_vector(&Vector3::new(1.0, 2.0, 3.0));
///
/// assert_relative_eq!(transformed_vector, Vector3::new(3.0, 2.0, -1.0), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use]
pub fn transform_vector(&self, v: &SVector<T, D>) -> SVector<T, D> {
self * v
}
/// Rotate the given point by the inverse of this rotation. This may be
/// cheaper than inverting the rotation and then transforming the given
/// point.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use std::f32;
/// # use nalgebra::{Point3, Rotation2, Rotation3, UnitQuaternion, Vector3};
/// let rot = Rotation3::new(Vector3::y() * f32::consts::FRAC_PI_2);
/// let transformed_point = rot.inverse_transform_point(&Point3::new(1.0, 2.0, 3.0));
///
/// assert_relative_eq!(transformed_point, Point3::new(-3.0, 2.0, 1.0), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use]
pub fn inverse_transform_point(&self, pt: &Point<T, D>) -> Point<T, D> {
Point::from(self.inverse_transform_vector(&pt.coords))
}
/// Rotate the given vector by the inverse of this rotation. This may be
/// cheaper than inverting the rotation and then transforming the given
/// vector.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use std::f32;
/// # use nalgebra::{Rotation2, Rotation3, UnitQuaternion, Vector3};
/// let rot = Rotation3::new(Vector3::y() * f32::consts::FRAC_PI_2);
/// let transformed_vector = rot.inverse_transform_vector(&Vector3::new(1.0, 2.0, 3.0));
///
/// assert_relative_eq!(transformed_vector, Vector3::new(-3.0, 2.0, 1.0), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use]
pub fn inverse_transform_vector(&self, v: &SVector<T, D>) -> SVector<T, D> {
self.matrix().tr_mul(v)
}
/// Rotate the given vector by the inverse of this rotation. This may be
/// cheaper than inverting the rotation and then transforming the given
/// vector.
///
/// # Example
/// ```
/// # #[macro_use] extern crate approx;
/// # use std::f32;
/// # use nalgebra::{Rotation2, Rotation3, UnitQuaternion, Vector3};
/// let rot = Rotation3::new(Vector3::z() * f32::consts::FRAC_PI_2);
/// let transformed_vector = rot.inverse_transform_unit_vector(&Vector3::x_axis());
///
/// assert_relative_eq!(transformed_vector, -Vector3::y_axis(), epsilon = 1.0e-6);
/// ```
#[inline]
#[must_use]
pub fn inverse_transform_unit_vector(&self, v: &Unit<SVector<T, D>>) -> Unit<SVector<T, D>> {
Unit::new_unchecked(self.inverse_transform_vector(&**v))
}
}
impl<T: Scalar + Eq, const D: usize> Eq for Rotation<T, D> {}
impl<T: Scalar + PartialEq, const D: usize> PartialEq for Rotation<T, D> {
#[inline]
fn eq(&self, right: &Self) -> bool {
self.matrix == right.matrix
}
}
impl<T, const D: usize> AbsDiffEq for Rotation<T, D>
where
T: Scalar + AbsDiffEq,
T::Epsilon: Clone,
{
type Epsilon = T::Epsilon;
#[inline]
fn default_epsilon() -> Self::Epsilon {
T::default_epsilon()
}
#[inline]
fn abs_diff_eq(&self, other: &Self, epsilon: Self::Epsilon) -> bool {
self.matrix.abs_diff_eq(&other.matrix, epsilon)
}
}
impl<T, const D: usize> RelativeEq for Rotation<T, D>
where
T: Scalar + RelativeEq,
T::Epsilon: Clone,
{
#[inline]
fn default_max_relative() -> Self::Epsilon {
T::default_max_relative()
}
#[inline]
fn relative_eq(
&self,
other: &Self,
epsilon: Self::Epsilon,
max_relative: Self::Epsilon,
) -> bool {
self.matrix
.relative_eq(&other.matrix, epsilon, max_relative)
}
}
impl<T, const D: usize> UlpsEq for Rotation<T, D>
where
T: Scalar + UlpsEq,
T::Epsilon: Clone,
{
#[inline]
fn default_max_ulps() -> u32 {
T::default_max_ulps()
}
#[inline]
fn ulps_eq(&self, other: &Self, epsilon: Self::Epsilon, max_ulps: u32) -> bool {
self.matrix.ulps_eq(&other.matrix, epsilon, max_ulps)
}
}
/*
*
* Display
*
*/
impl<T, const D: usize> fmt::Display for Rotation<T, D>
where
T: RealField + fmt::Display,
{
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
let precision = f.precision().unwrap_or(3);
writeln!(f, "Rotation matrix {{")?;
write!(f, "{:.*}", precision, self.matrix)?;
writeln!(f, "}}")
}
}
// // /*
// // *
// // * Absolute
// // *
// // */
// // impl<T: Absolute> Absolute for $t<T> {
// // type AbsoluteValue = $submatrix<T::AbsoluteValue>;
// //
// // #[inline]
// // fn abs(m: &$t<T>) -> $submatrix<T::AbsoluteValue> {
// // Absolute::abs(&m.submatrix)
// // }
// // }