Statistics Toolkit

A personal repository of Applied Statistics workflows that combine clean implementation with explanatory depth.

Each notebook covers a complete workflow — from setup to summary — while staying readable to both technical and non-technical users. Theory is seamlessly integrated to strengthen understanding.

🧠 What's Inside

1. A/B Testing (Notebook)

• Experiment setup with config-driven control
• Randomization logic (simple, cluster, match-pair)
• AA testing and diagnostics
• Sample size & power analysis

2. Hypothesis Testing (Notebook)

• Parametric and non-parametric test selection
• Normality and variance checks
• Paired vs unpaired logic
• Significance interpretation and output labeling

3. Causal Inference (Notebook)

• Estimating treatment effects from observational data
• Potential outcomes framework (ATE, ATT)
• Propensity score matching, stratification, weighting
• Supports integration with experimental pipelines

4. Statistical Distributions (Notebook)

• Continuous: Uniform, Normal, Exponential, Chi-Square, Student t
• Discrete: Poisson, Bernoulli, Binomial
• Visualizations, sampling, parameter estimation

5. Statistical Paradoxes (Notebook)

• Simpson’s Paradox
• Will Rogers Phenomenon

6. Statistical Similarity Metrics (Notebook)

• Euclidean, Manhattan, Minkowski, Mahalanobis distances
• Cosine similarity/distance, Jaccard index, Hamming distance
• Pearson and Spearman correlation

7. Theoretical Foundations (Notebook)

• Law of Large Numbers (with visual convergence)
• Central Limit Theorem (sample mean simulation)
• Bayes’ Theorem (belief updating using evidence)
• Chebyshev’s Inequality (distribution-free bounds on data spread)

---