Environment Setup¶

Quick Start with Google Colab¶

The fastest way to start is Google Colab — no installation required. All libraries are pre-installed.

Local Installation¶

Step 1: Install Python¶

Download Python 3.10+ from python.org.

Step 2: Create a Virtual Environment¶

python -m venv ml-env
# Windows
ml-env\\Scripts\\activate
# Mac/Linux
source ml-env/bin/activate

Step 3: Install Libraries¶

pip install pandas numpy matplotlib seaborn scikit-learn
pip install xgboost lightgbm
pip install shap lime
pip install optuna
pip install statsmodels prophet
pip install jupyter

Step 4: Launch Jupyter¶

jupyter notebook

Library Usage Map¶

Library	Used For	Topics
`pandas`	Data loading, manipulation, cleaning	All
`numpy`	Numerical operations	All
`matplotlib` / `seaborn`	Visualisation	All
`scikit-learn`	ML algorithms, preprocessing, evaluation	1–5, 7
`xgboost` / `lightgbm`	Gradient boosting	3, 4
`statsmodels`	Time series analysis, statistical tests	6
`prophet`	Automated time series forecasting	6
`shap` / `lime`	Model interpretability	8
`optuna`	Bayesian hyperparameter optimisation	7

Workplace Tip

Keep a requirements.txt in your project so colleagues can reproduce your environment: pip freeze > requirements.txt