MN_module8

Numerical Methods for Nonlinear Physics - Module 8. Pisa University

This repository contains code and resources developed for Module 8 of the course "Numerical Methods for Physics". The project focuses on statistical analysis, mutual information estimation, and synthetic data generation.

---

Features

• Data Generation: Tools for generating univariate and multivariate synthetic datasets.
• Mutual Information Calculation: Implementation of algorithms to calculate mutual information (MI) with k-nearest neighbors.
• Data Checking Utilities: Scripts for validating and visualizing datasets.

---

Project Structure

Repository Structure: MN_module8

Top-Level Directories

• core/ : Scripts for Mutual Information (MI) calculation.
• sampling/ : Scripts for synthetic data generation.
• utils/ : Utility scripts and helper functions.
• tests/ : Unit tests for core functionalities.
• old_repo_architecture/ : Archived repository structure files (until 12/12/2024).

---

Directory Breakdown

core/

Scripts for computing Mutual Information (MI):

• calculate_mi_algorithm1.py : MI calculation using Algorithm 1 from Kraskov et al.
• calculate_mi_algorithm1_entropies_sum.py : Summing entropies for MI estimation.
• calculate_mi_algorithm2.py : MI calculation using Algorithm 2 from Kraskov et al.

sampling/

Scripts for generating synthetic datasets:

• generate_data.py : Main script for data generation.
• generate_multivariate_data.py : Generates bivariate datasets.
• generate_univariate_data.py : Generates a dataset composed by two univariate correlated series.

utils/

Utility scripts to support core tasks:

• config_utils.py: Handles configuration files
• decorators.py:
• io_utils.py: handles reading, writing, managing data files efficiently and manages input/output operations.
• math_utils.py: Contains helper functions for mathematical operations
• mutual_information_utils.py: Utility functions to support MI computation algorithms
• plot_utils.py : Functions for visualizing data (e.g., line charts, scatter plots, histograms).
• pre_processing_utils.py: Pre-processes data (normalization, cleaning) before analysis.

tests/

Unit tests to ensure code reliability:

Other Files

• calculate_mi_algorithm1.py: Top-level MI script
• calculate_mi_algorithm1_entropies_sum.py: Top-level MI script
• calculate_mi_algorithm2.py: Top-level MI script
• checking_data.py : Validates and checks data integrity.
• generate_data.py : Top-level data generation script
• generate_multivariate_data.py : Top-level data generation script
• generate_univariate_data.py : Top-level data generation script
• config.json: Configuration file
• .gitignore : Specifies files and directories ignored by Git.
• LICENSE : Project licensing information (GPL-3.0).
• README.md : Main documentation for the repository.

---