Advanced Statistical Computing with Python

This site presents the material from Advanced Statistical Computing by Roger Peng, originally written in R. Best efforts have been made to faithfully reproduce it in Python using NumPy, SciPy, and Matplotlib.

Chapters

1. Introduction — Linear models, optimization transfer, textbooks vs. computers
2. Solving Nonlinear Equations — Bisection, rates of convergence, functional iteration, Newton's method
3. General Optimization — Steepest descent, Newton direction, quasi-Newton, conjugate gradient, coordinate descent
4. The EM Algorithm — Exponential families, canonical examples, minorizing functions, missing information, acceleration
5. Integration — Laplace approximation
6. Independent Monte Carlo — Random number generation, non-uniform sampling, rejection sampling, importance sampling
7. Markov Chain Monte Carlo — Metropolis-Hastings, Gibbs sampler, convergence monitoring, simulated annealing