Overwiew

MatrixLibs is C++11 meta-programming, header-only, no-dynamic memory allocations library for matrix computation, similar to matlab.

Algorithms already implemented:

• Sum
• Substraction
• Multiply
• Division
• Absolut
• Power
• Transpose
• Inverse
• Strassen
• Determinant
• LU decomposition
• Cholesky decomposition
• Concatenate
• Convolution
• Histogram
• Filters: mean, meadina, variance, standard deviation
• Swap rows, cols

Usage

• include matrix.hpp
• enable C++11 or higher
• have fun

Matrix object uses std::array from C++11, so the explicity type and the size of matrix is always necessary

constexpr std::size_t n = 2;
Matrix<float, n, n> A
{ 1, 2, 
  3, 4 };

auto B = <float, n, n>eye();

auto C = A * B:

Dependiences

The google test and google benchmark framework is used in correct and performance tests

• https://github.com/google/googletest
• https://github.com/google/benchmark

Compiling

Compile like most of applications with cmake

$ mkdir build && cd build
$ cmake ..
$ make

If you don't want to make the tests, you can use this cmake options

$ cmake -DBUILD_TESTS=OFF

or switch off benchmark tests

$ cmake -DBUILD_BENCH=OFF

If you want to debug the memory leaks, you can set up the address sanitizer option

$ cmake -DSANITIZER=ON