The "NumPy" Approach
summary
Shows a concrete performance claim: rewriting a binary genetic algorithm from Python lists and explicit loops into NumPy arrays and vectorized ops cuts runtime (example: naive Python matrix multiply 14.2s vs NumPy 164ms; full GA runs drop from ~2–2.6s to ~1s for larger populations). Implements ndarray masking, flattened index put() updates and XOR bit flips, generates populations with NumPy random, performs crossover via elementwise masks, and reports correct convergence to all-ones.
AI-written summary · about
chapters (10)
- 0:00NumPy approach
- 1:24Pure Python vector ops
- 4:04Genetic algorithm design
- 16:05Python lists limits
- 18:30NumPy machine types
- 20:55Algorithm with NumPy arrays
- 25:31Evolution in 3D NumPy
- 33:18Optimizing bit flips
- 34:44Performance comparison
- 35:08Conclusion
related talks
34mNumPy: vectorize your brain
46mTight Genes: Intro to Genetic Algorithms
29mVectorize using linear algebra and NumPy to make your Python code fast
34mAccelerating GPU Algorithms in pure Python
27mAPL-Inspired Techniques for Advanced NumPy
01h 27mMaking computation easier with cool Numpy tricks
38mVectorize all the things! Using linear algebra and NumPy to make your Python code lightning fast
01h 31mMaking Computation Easier with Cool Numpy Tricks