VMATH: A header only library for numpy like operation on C++ vector.

The main goal of this library is to offer a light-weight tool which mimics NUMPY like operation on C++ vector. For simple addition and subtraction, we always need to write for-loops in C++. It would be nice to do these operation like NUMPY. This library also contains some useful math operations, e.g. matmul, cross_product, dot_product, transpose etc.

Installation

Copy header file in your source directory and include into your cpp file.

Usages:

1. Overloading operators

This library contains various operator overloading to avoid writing explicit for-loop,

Example-1: Vector addition

#include "vmath.h" 

// Without vmath
for (int i=0; i<n;++i){
   A[i] = B[i] + C; 
}
// With vmath
A = B + C;

Example-2: Overloading console output "<<"

//Without vmath
for (int i=0;i<n;++i)
    std::cout<< A[i]<< " ";
std::endl; 

//With vmath 
std::cout<< A << std::endl;

Following operators can be overloaded using vmath.h

Addition: +, +=
Subtraction: -, -=
Multiplication: *, *=
Division: /, /=

2. Vector initialization like Numpy

// Get a vector/matrix initialized with 1 and size of n;
std::vector<float> vec = vmath::ones<float>(n); 
vmath::matrix<float> mat = vmath::ones<float>(n1,n2); 

// Get vector/matrix  with randomly initialized  values (0.0 to 1.0)
std::vector<float> vec= vmath::random<float>(n); 
vmath::matrix<float> mat= vmath::random<float>(n1,n2); 

// Get a vector/matrix initialized with 0 and size of n;
std::vector<float> vec = vmath::zeros<float>(n); 
vmath::matrix<float> mat = vmath::zeros<float>(n1,n2);

3. Vector cross_product, dot_product, sum, argmin, argmax

vector<float> u = {1,2,3};
vector<int> v = {3,2,1};


//cross_product
std::cout << vmath::cross_product(v,u) <<std::endl;

//dot_product
std::cout << vmath::dot_product(v,u) <<std::endl;

//sum  
std::cout << vmath::sum(u) << std::endl; 

//argmin 
std::cout << vmath::argmin(v) <<std::endl;

//argmax
std::cout << vmath::argmax(v) <<std::endl;

4. Matrix multiplication/transpose

// Defining matrix     
vmath::matrix<float> mat0=vmath::random<float>(5000,300);
vmath::matrix<float> mat1=vmath::random<float>(300,600);

// Multiply    
vmath::matrix<float> reslt= vmath::matmul(mat0,mat1);

// Transpose 
vmath::matrix<float> reslt= vmath::transpose(mat0);