My-Linear-Algebra-CPP

Simple Matrix Class implementation in C++ using Cuda GPU Programming.

How it works ?

Instantiate a new Matrix

/* Random Matrix of size 10, 10 */
Matrix<float> A = Matrix<float>(10, 10);

/* Empty matrix */
Matrix<double> B = Matrix<double>();

/* Random Matrix of size 10, 10 with values between min & max */
Matrix<float> C = Matrix<float>(10, 10, -0.5f, 0.5f);

/* Matrix of size 2, 3 with initializer list */
Matrix<int> D = Matrix<int>({{1, 2, 3}, {4, 5, 6}});

Bastic operations

/* Basic Operations between Matrix */
Matrix<float> E = A + A;
Matrix<float> G = A - A;
Matrix<float> F = A * A;
Matrix<float> G = A / A;
bool eq = (E == G);

/* Basic Operations between Matrix & Number */
Matrix<float> H = 2.f + A;
Matrix<float> I = 2.f - A;
Matrix<float> J = 2.f * A;
Matrix<float> K = 2.f / A;

Methods

/* Element wise operation */
Matrix<float> L = A.exponential();
Matrix<float> M = A.power(2);

/* Other methods */
Matrix<float> N = A.getLine(5);
Matrix<float> O = A.getRow(5);
Matrix<float> P = A.transpose();
Matrix<float> P = A.apply([](float x) { return 2.f * x; });

Benchmarks