Matrix for C++

A tiny C++20 general-purpose library for dense, fixed-size and dynamically-sized matrices with no dependencies (other than std). It makes use of C++20s concepts to enable special features such as norm for vectors and many more.

View the full documentation at https://mc-zen.github.io/Matrix/

Overview

Features:

• intuitive syntax
• low-overhead
• header-only and single-header
• fully constexpr (except where math functions like sqrt are used).
• STL-oriented
• Statically-sized have static memory (e.g. suited for audio development).
• Dynamically-sized matrices are now available as well.
• simple software design in comparison to some other matrix libs, e.g. no CRTP, virtual functions, ...
• fully tested. Tests can be run with Catch2.

Limitations:

• only C++20 (and up) because of heavy use of concepts and constexpr
• this is not a linear algebra library - at the moment the library provides no linear algebra algorithms such as decomposition and solvers (exception: determinant and inverse of 2×2 and 3×3 matrices).

Supported Platforms

Since this repository uses C++20 language features, a fairly new C++ compiler is necessary.

Platform	Compiler	Minimum version
Windows	MSVC	14.30
Linux	GCC	13
	Clang	15
MacOS	AppleClang	15

Classes

Class	Definition
Matrix<T,m,n>	Main class for m×n matrices of type T
Matrix_view<T>	Matrix view class for viewing or modifying parts of a matrix like a block, row or column.
Matrix_iterator<T>	Random access iterator for iterating through a contiguous array in memory (matrix data).
Matrix_stride_iterator<T>	Random access iterator for iterating i.e. vertically through a contiguous array in memory.
Matrix_block_iterator<T>	Bi-directional iterator for iterating through a matrix block within a bigger matrix.