🧮 TensorOpsInPython

A Python repository demonstrating tensor operations implemented from scratch, compared with NumPy's efficient implementations.

📝 Description

TensorOpsInPython provides educational examples of fundamental tensor operations implemented using naive Python loops alongside their NumPy equivalents. This repository serves as a learning resource for understanding the underlying mechanics of tensor operations commonly used in machine learning and deep learning applications.

✨ Features

• Basic Tensor Operations:

  • ReLU activation function
  • Addition, subtraction, and multiplication of tensors
  • Broadcasting (adding a matrix and a vector)

• Advanced Tensor Operations:

  • Vector dot product
  • Matrix-vector dot product (two implementations)
  • Matrix-matrix dot product

• Tensor Transformations:

  • Reshaping tensors
  • Transposing tensors (exchanging rows and columns)

• Comparative Learning:

  • Each operation is implemented both with naive Python loops and using NumPy's optimized functions
  • Output comparison between custom implementations and NumPy equivalents

🔧 Prerequisites

• Python 3.x
• NumPy library

🚀 Usage

Run the main script to see all tensor operations in action:

python main.py

Example Code Snippets

Implementing a Naive ReLU Function

@staticmethod
def naive_relu(x):
    assert len(x.shape) == 2
    x = x.copy()
    for i in range(x.shape[0]):
        for j in range(x.shape[1]):
            x[i, j] = max(x[i, j], 0)
    return x

Matrix-Vector Dot Product

@staticmethod
def naive_matrix_vector_dot(x, y):
    assert len(x.shape) == 2  # Numpy matrix
    assert len(y.shape) == 1  # Numpy vector
    assert x.shape[1] == y.shape[0]

    z = np.zeros(x.shape[0])
    for i in range(x.shape[0]):
        for j in range(x.shape[1]):
            z[i] += x[i, j] * y[j]
    return z

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.