Neupy

Overview

Neupy is a basic implementation of the fundamental features of the NumPy library written in C. It provides multi-dimensional array support and operations commonly used in scientific computing, machine learning, and data manipulation. This project simulates the core array-based operations such as array creation, mathematical operations, reshaping, and matrix multiplication, offering a foundation for array manipulation in a style similar to Python’s NumPy.

Features

• Multi-dimensional arrays: Support for creating arrays with arbitrary dimensions.
• Mathematical operations: Add, subtract, multiply, and divide arrays element-wise.
• Matrix operations: Matrix multiplication using traditional and Strassen algorithms.
• Reshaping & Transposing: Reshape arrays into different dimensions or transpose matrices.
• Random Arrays: Generate arrays with random values within a specified range.
• Trigonometric & Exponential Functions: Compute element-wise trigonometric functions like sine, cosine, and tangent, as well as power, square root, and exponentiation functions.

Categories of Functions

1. Array Creation Functions

• create_NeuArray: Create an empty array with a given shape.
• zeros: Create an array filled with zeros.
• ones: Create an array filled with ones.
• eye: Create an identity matrix.
• randint: Create an array filled with random integers.

2. Array Inspection Functions

• print_NeuArray: Print the contents and shape of an array.
• len: Calculate the total number of elements in an array.
• compare_NeuArray: Compare two arrays element-wise.
• max: Find the maximum value in the array.
• min: Find the minimum value in the array.
• mean: Calculate the mean of the array.

3. Mathematical Functions

• add: Element-wise addition of two arrays.
• subtract: Element-wise subtraction of two arrays.
• multiply: Element-wise multiplication of two arrays.
• divide: Element-wise division of two arrays.
• neg: Negate all elements in an array.
• sqrt_array: Compute the square root of each element in an array.
• exp_array: Compute the exponential (e^x) of each element in an array.
• power: Raise each element in the array to the given power.

4. Trigonometric Functions

• sin: Compute the sine of each element in an array.
• cos: Compute the cosine of each element in an array.
• tan: Compute the tangent of each element in an array.

5. Linear Algebra Functions

• matrix_multiply: Perform matrix multiplication.
• strassen_multiply: Perform matrix multiplication using the Strassen algorithm.
• transpose: Transpose a 2D matrix.

6. Array Reshaping Functions

• reshape: Change the shape of an array while maintaining the number of elements.
• free_NeuArray: Free memory allocated to the array.

How to Use

Prerequisites

• GCC (or any other C compiler)
• Standard C Libraries (stdlib.h, stdio.h, etc.)

Compilation

To compile the project, run the following command:

gcc -o neuarray main.c -lm

Here, main.c is the source file where you will be using the NeuArray library. The -lm flag links the math library for functions like pow, sqrt, etc.

Using NeuArray in Your Project

1. Include the necessary headers:

#include "neuarray.h"

2. Create arrays and perform operations:

int shape[] = {3, 3};
NeuArray* arr1 = ones(shape, 2);
NeuArray* arr2 = randint(1, 10, shape, 2);

// Add the arrays
NeuArray* sum = add(arr1, arr2);

// Print the result
print_NeuArray(sum);

// Free memory
free_NeuArray(arr1);
free_NeuArray(arr2);
free_NeuArray(sum);

Example Code

#include <stdio.h>
#include "neuarray.h"

int main() {
    int shape[] = {2, 2};

    // Create two arrays with random integers
    NeuArray* A = randint(0, 10, shape, 2);
    NeuArray* B = randint(0, 10, shape, 2);

    // Perform matrix multiplication
    NeuArray* C = matrix_multiply(A, B);

    // Print the arrays and result
    printf("Matrix A:\n");
    print_NeuArray(A);

    printf("Matrix B:\n");
    print_NeuArray(B);

    printf("Matrix A * B:\n");
    print_NeuArray(C);

    // Free allocated memory
    free_NeuArray(A);
    free_NeuArray(B);
    free_NeuArray(C);

    return 0;
}