Trait cgmath::Matrix

source ·

pub trait Matrix: VectorSpacewhere
    Self::Scalar: BaseFloat,
    Self: Index<usize, Output = Self::Column> + IndexMut<usize, Output = Self::Column> + ApproxEq<Epsilon = <Self as VectorSpace>::Scalar>,{
    type Row: VectorSpace<Scalar = Self::Scalar> + Array<Element = Self::Scalar>;
    type Column: VectorSpace<Scalar = Self::Scalar> + Array<Element = Self::Scalar>;
    type Transpose: Matrix<Scalar = Self::Scalar, Row = Self::Column, Column = Self::Row>;

    // Required methods
    fn row(&self, r: usize) -> Self::Row;
    fn swap_rows(&mut self, a: usize, b: usize);
    fn swap_columns(&mut self, a: usize, b: usize);
    fn swap_elements(&mut self, a: (usize, usize), b: (usize, usize));
    fn transpose(&self) -> Self::Transpose;

    // Provided methods
    fn as_ptr(&self) -> *const Self::Scalar { ... }
    fn as_mut_ptr(&mut self) -> *mut Self::Scalar { ... }
    fn replace_col(&mut self, c: usize, src: Self::Column) -> Self::Column { ... }
}

Expand description

A column-major matrix of arbitrary dimensions.

Because this is constrained to the VectorSpace trait, this means that following operators are required to be implemented:

Matrix addition:

Add<Output = Self>
Sub<Output = Self>
Neg<Output = Self>

Scalar multiplication:

Mul<Self::Scalar, Output = Self>
Div<Self::Scalar, Output = Self>
Rem<Self::Scalar, Output = Self>

Note that matrix multiplication is not required for implementors of this trait. This is due to the complexities of implementing these operators with Rust’s current type system. For the multiplication of square matrices, see SquareMatrix.