Microbenchmarks showing relative performance of different Python functions/patterns.

These are designed to answer the question "Is code snippet X faster than code snippet Y?". Several microbenchmarks have baked-in assumptions about data sizes and access patterns, so be careful not to overgeneralize the results.