Struct ultraviolet::vec::DVec3x4
source · #[repr(C)]pub struct DVec3x4 {
pub x: f64x4,
pub y: f64x4,
pub z: f64x4,
}
Expand description
A set of three coordinates which may be interpreted as a point or vector in 3d space, or as a homogeneous 2d vector or point.
Generally this distinction between a point and vector is more of a pain than it is worth to distinguish on a type level, however when converting to and from homogeneous coordinates it is quite important.
Fields§
§x: f64x4
§y: f64x4
§z: f64x4
Implementations§
source§impl DVec3x4
impl DVec3x4
pub const fn new(x: f64x4, y: f64x4, z: f64x4) -> Self
pub const fn broadcast(val: f64x4) -> Self
pub fn unit_x() -> Self
pub fn unit_y() -> Self
pub fn unit_z() -> Self
sourcepub fn into_homogeneous_point(self) -> DVec4x4
pub fn into_homogeneous_point(self) -> DVec4x4
Create a homogeneous 3d point from this vector interpreted as a point, meaning the homogeneous component will start with a value of 1.0.
sourcepub fn into_homogeneous_vector(self) -> DVec4x4
pub fn into_homogeneous_vector(self) -> DVec4x4
Create a homogeneous 3d vector from this vector, meaning the homogeneous component will always have a value of 0.0.
sourcepub fn from_homogeneous_point(v: DVec4x4) -> Self
pub fn from_homogeneous_point(v: DVec4x4) -> Self
Create a 3d point from a homogeneous 3d point, performing division by the homogeneous component. This should not be used for homogeneous 3d vectors, which will have 0 as their homogeneous component.
sourcepub fn from_homogeneous_vector(v: DVec4x4) -> Self
pub fn from_homogeneous_vector(v: DVec4x4) -> Self
Create a 3d vector from homogeneous 2d vector, which simply discards the homogeneous component.
pub fn dot(&self, other: DVec3x4) -> f64x4
sourcepub fn wedge(&self, other: DVec3x4) -> DBivec3x4
pub fn wedge(&self, other: DVec3x4) -> DBivec3x4
The wedge (aka exterior) product of two vectors.
This operation results in a bivector, which represents
the plane parallel to the two vectors, and which has a
‘oriented area’ equal to the parallelogram created by extending
the two vectors, oriented such that the positive direction is the
one which would move self
closer to other
.
sourcepub fn geom(&self, other: DVec3x4) -> DRotor3x4
pub fn geom(&self, other: DVec3x4) -> DRotor3x4
The geometric product of this and another vector, which is defined as the sum of the dot product and the wedge product.
This operation results in a ‘rotor’, named as such as it may define
a rotation. The rotor which results from the geometric product
will rotate in the plane parallel to the two vectors, by twice the angle between
them and in the opposite direction (i.e. it will rotate in the direction that would
bring other
towards self
, and rotate in that direction by twice the angle between them).
pub fn rotate_by(&mut self, rotor: DRotor3x4)
pub fn rotated_by(self, rotor: DRotor3x4) -> Self
pub fn cross(&self, other: DVec3x4) -> Self
pub fn reflect(&mut self, normal: DVec3x4)
pub fn reflected(&self, normal: DVec3x4) -> Self
pub fn mag_sq(&self) -> f64x4
pub fn mag(&self) -> f64x4
pub fn normalize(&mut self)
pub fn normalized(&self) -> Self
sourcepub fn normalize_homogeneous_point(&mut self)
pub fn normalize_homogeneous_point(&mut self)
Normalize self
in-place by interpreting it as a homogeneous point, i.e.
scaling the vector to ensure the homogeneous component has length 1.
sourcepub fn normalized_homogeneous_point(&self) -> Self
pub fn normalized_homogeneous_point(&self) -> Self
Normalize self
by interpreting it as a homogeneous point, i.e.
scaling the vector to ensure the homogeneous component has length 1.
sourcepub fn truncated(&self) -> DVec2x4
pub fn truncated(&self) -> DVec2x4
Convert self
into a Vec2 by simply removing its z
component.
pub fn mul_add(&self, mul: DVec3x4, add: DVec3x4) -> Self
pub fn abs(&self) -> Self
pub fn clamp(&mut self, min: Self, max: Self)
pub fn clamped(self, min: Self, max: Self) -> Self
pub fn map<F>(&self, f: F) -> Selfwhere F: FnMut(f64x4) -> f64x4,
pub fn apply<F>(&mut self, f: F)where F: FnMut(f64x4) -> f64x4,
pub fn max_by_component(self, other: Self) -> Self
pub fn min_by_component(self, other: Self) -> Self
pub fn component_max(&self) -> f64x4
pub fn component_min(&self) -> f64x4
pub fn zero() -> Self
pub fn one() -> Self
pub const fn xy(&self) -> DVec2x4
pub fn xyzw(&self) -> DVec4x4
sourcepub fn layout() -> Layout
pub fn layout() -> Layout
Get the core::alloc::Layout
of Self
sourcepub fn as_array(&self) -> &[f64x4; 3]
pub fn as_array(&self) -> &[f64x4; 3]
Interpret self
as a statically-sized array of its base numeric type
sourcepub fn as_mut_array(&mut self) -> &mut [f64x4; 3]
pub fn as_mut_array(&mut self) -> &mut [f64x4; 3]
Interpret self
as a statically-sized array of its base numeric type
sourcepub fn as_mut_slice(&mut self) -> &mut [f64x4]
pub fn as_mut_slice(&mut self) -> &mut [f64x4]
Interpret self
as a slice of its base numeric type
pub fn as_byte_slice(&self) -> &[u8] ⓘ
pub fn as_mut_byte_slice(&mut self) -> &mut [u8] ⓘ
sourcepub const fn as_ptr(&self) -> *const f64x4
pub const fn as_ptr(&self) -> *const f64x4
Returns a constant unsafe pointer to the underlying data in the underlying type. This function is safe because all types here are repr(C) and can be represented as their underlying type.
Safety
It is up to the caller to correctly use this pointer and its bounds.
sourcepub fn as_mut_ptr(&mut self) -> *mut f64x4
pub fn as_mut_ptr(&mut self) -> *mut f64x4
Returns a mutable unsafe pointer to the underlying data in the underlying type. This function is safe because all types here are repr(C) and can be represented as their underlying type.
Safety
It is up to the caller to correctly use this pointer and its bounds.
source§impl DVec3x4
impl DVec3x4
pub fn new_splat(x: f64, y: f64, z: f64) -> Self
pub fn splat(vec: DVec3) -> Self
sourcepub fn blend(mask: m64x4, tru: Self, fals: Self) -> Self
pub fn blend(mask: m64x4, tru: Self, fals: Self) -> Self
Blend two vectors together lanewise using mask
as a mask.
This is essentially a bitwise blend operation, such that any point where
there is a 1 bit in mask
, the output will put the bit from tru
, while
where there is a 0 bit in mask
, the output will put the bit from fals
pub fn refract(&mut self, normal: Self, eta: f64x4)
pub fn refracted(&self, normal: Self, eta: f64x4) -> Self
Trait Implementations§
source§impl AddAssign<DVec3x4> for DVec3x4
impl AddAssign<DVec3x4> for DVec3x4
source§fn add_assign(&mut self, rhs: DVec3x4)
fn add_assign(&mut self, rhs: DVec3x4)
+=
operation. Read moresource§impl DivAssign<DVec3x4> for DVec3x4
impl DivAssign<DVec3x4> for DVec3x4
source§fn div_assign(&mut self, rhs: DVec3x4)
fn div_assign(&mut self, rhs: DVec3x4)
/=
operation. Read moresource§impl DivAssign<f64x4> for DVec3x4
impl DivAssign<f64x4> for DVec3x4
source§fn div_assign(&mut self, rhs: f64x4)
fn div_assign(&mut self, rhs: f64x4)
/=
operation. Read moresource§impl Lerp<f64x4> for DVec3x4
impl Lerp<f64x4> for DVec3x4
source§fn lerp(&self, end: Self, t: f64x4) -> Self
fn lerp(&self, end: Self, t: f64x4) -> Self
Linearly interpolate between self
and end
by t
between 0.0 and 1.0.
i.e. (1.0 - t) * self + (t) * end
.
For interpolating Rotor
s with linear interpolation, you almost certainly
want to normalize the returned Rotor
. For example,
let interpolated_rotor = rotor1.lerp(rotor2, 0.5).normalized();
For most cases (especially where performance is the primary concern, like in
animation interpolation for games, this ‘normalized lerp’ or ‘nlerp’ is probably
what you want to use. However, there are situations in which you really want
the interpolation between two Rotor
s to be of constant angular velocity. In this
case, check out Slerp
.
source§impl Mul<DVec3x4> for DIsometry3x4
impl Mul<DVec3x4> for DIsometry3x4
source§impl Mul<DVec3x4> for DSimilarity3x4
impl Mul<DVec3x4> for DSimilarity3x4
source§impl MulAssign<DVec3x4> for DVec3x4
impl MulAssign<DVec3x4> for DVec3x4
source§fn mul_assign(&mut self, rhs: DVec3x4)
fn mul_assign(&mut self, rhs: DVec3x4)
*=
operation. Read moresource§impl MulAssign<f64x4> for DVec3x4
impl MulAssign<f64x4> for DVec3x4
source§fn mul_assign(&mut self, rhs: f64x4)
fn mul_assign(&mut self, rhs: f64x4)
*=
operation. Read moresource§impl PartialEq<DVec3x4> for DVec3x4
impl PartialEq<DVec3x4> for DVec3x4
source§impl Slerp<f64x4> for DVec3x4
impl Slerp<f64x4> for DVec3x4
source§fn slerp(&self, end: Self, t: f64x4) -> Self
fn slerp(&self, end: Self, t: f64x4) -> Self
Spherical-linear interpolation between self
and end
based on t
from 0.0 to 1.0.
self
and end
should both be normalized or something bad will happen!
The implementation for SIMD types also requires that the two things being interpolated between are not exactly aligned, or else the result is undefined.
Basically, interpolation that maintains a constant angular velocity
from one orientation on a unit hypersphere to another. This is sorta the “high quality” interpolation
for Rotor
s, and it can also be used to interpolate other things, one example being interpolation of
3d normal vectors.
Note that you should often normalize the result returned by this operation, when working with Rotor
s, etc!
source§impl SubAssign<DVec3x4> for DVec3x4
impl SubAssign<DVec3x4> for DVec3x4
source§fn sub_assign(&mut self, rhs: DVec3x4)
fn sub_assign(&mut self, rhs: DVec3x4)
-=
operation. Read more