dragon-ml-toolbox 18.10.1

Complete pipelines and helper tools for data science and machine learning projects.

dragon-ml-toolbox

A collection of machine learning pipelines and utilities, structured as modular packages for easy reuse and installation. This package has no base dependencies, allowing for lightweight and customized virtual environments.

Features:

• Modular scripts for data exploration, logging, machine learning, and more.
• Designed for seamless integration as a Git submodule or installable Python package.

Installation

Python 3.12

Via PyPI

Install the latest stable release from PyPI:

pip install dragon-ml-toolbox

Via conda-forge

Install from the conda-forge channel:

conda install -c conda-forge dragon-ml-toolbox

Via GitHub (Editable)

Clone the repository and install in editable mode:

git clone https://github.com/DrAg0n-BoRn/ML_tools.git
cd ML_tools
pip install -e .

Modular Installation

This toolbox is designed as a collection of mutually exclusive environments due to conflicting core dependencies, except APP bundlers (PyInstaller/Nuitka).

• Rule: Create a fresh virtual environment for each module to use.

📦 Core Machine Learning Toolbox [ML]

Installs a comprehensive set of tools for typical data science workflows, including data manipulation, modeling, and evaluation using PyTorch.

➡️ Prerequisite: PyTorch required. Follow the official instructions: PyTorch website

pip install "dragon-ml-toolbox[ML]"

Modules:

constants
custom_logger
data_exploration
ETL_cleaning
ETL_engineering
math_utilities
ML_callbacks
ML_configuration
ML_datasetmaster
ML_evaluation_captum
ML_evaluation_multi
ML_evaluation
ML_inference
ML_models
ML_models_advanced
ML_optimization_pareto
ML_optimization
ML_scaler
ML_sequence_datasetmaster
ML_sequence_evaluation
ML_sequence_inference
ML_sequence_models
ML_trainer
ML_utilities
ML_vision_datasetmaster
ML_vision_evaluation
ML_vision_inference
ML_vision_models
ML_vision_transformers
optimization_tools
path_manager
schema
serde
SQL
utilities

Wrappers for some pytorch_tabular models are available:

pip install "dragon-ml-toolbox[ML,py-tab]"

# Extra Modules:
ML_models_pytab
ML_configuration_pytab

---

🌳 Ensemble Learning [ensemble]

Comprehensive set of tools for typical data science workflows focused on XGBoost and LightGBM.

Note: Optimizes for NumPy >= 2.0.

pip install "dragon-ml-toolbox[ensemble]"

Modules:

constants
custom_logger
data_exploration
ensemble_evaluation
ensemble_inference
ensemble_learning
ETL_cleaning
ETL_engineering
math_utilities
optimization_tools
path_manager
PSO_optimization
schema
serde
SQL
utilities

---

🔬 MICE Imputation and Variance Inflation Factor [mice]

Utilities for advanced data cleaning and statistical checks. Features Multiple Imputation by Chained Equations (MICE) for handling missing data and Variance Inflation Factor (VIF) analysis to detect multicollinearity in features.

pip install "dragon-ml-toolbox[mice]"

Modules:

constants
custom_logger
math_utilities
MICE_imputation
serde
VIF_factor
path_manager
utilities

---

📋 Excel File Handling [excel]

Installs dependencies required to process and handle .xlsx or .xls files.

pip install "dragon-ml-toolbox[excel]"

Modules:

custom_logger
excel_handler
path_manager

---

🎰 GUI for Boosting Algorithms (XGBoost, LightGBM) [gui-boost]

GUI tools compatible with XGBoost and LightGBM models used for inference.

pip install "dragon-ml-toolbox[gui-boost]"

Modules:

constants
custom_logger
GUI_tools
ensemble_inference
path_manager
schema
serde

---

🤖 GUI for PyTorch Models [gui-torch]

GUI tools compatible with PyTorch models used for inference.

pip install "dragon-ml-toolbox[gui-torch]"

Modules:

constants
custom_logger
GUI_tools
ML_models
ML_models_advanced
ML_sequence_models
ML_vision_models # Requires: torchvision and Pillow
ML_inference
ML_sequence_inference
ML_vision_inference # Requires: torchvision and Pillow
ML_vision_transformers # Requires: torchvision and Pillow
ML_scaler
path_manager
schema

---

⚒️ APP bundlers

Dependencies required to compile applications, inference scripts, or GUIs into standalone executables (.exe or binary) for distribution. Choose your preferred backend:

pip install "dragon-ml-toolbox[pyinstaller]"

pip install "dragon-ml-toolbox[nuitka]"

Usage

After installation, import modules like this:

from ml_tools.serde import serialize_object, deserialize_object
from ml_tools.custom_logger import train_logger