python-performance-optimization

Profiling Python code identifies which functions consume the most CPU or memory time. Use cProfile for function-level CPU profiling, line_profiler for line-by-line analysis, and memory_profiler to track memory consumption. Start with cProfile to locate slow sections, then zoom in with line_profiler for detailed optimization targets.

Several profiling tools target different bottlenecks: cProfile and py-spy measure CPU usage, memory_profiler tracks memory allocation, and asyncio patterns optimize I/O-bound workloads while multiprocessing accelerates CPU-bound tasks. Choose based on whether your bottleneck is CPU, memory, or I/O.

Memory optimization involves profiling with memory_profiler to identify allocation hotspots, restructuring data structures to use less overhead, and applying lazy-loading patterns. The skill covers patterns for data-intensive workloads where memory footprint directly impacts performance.

Yes. Asyncio patterns optimize I/O-bound operations through concurrent coroutines, while multiprocessing parallelizes CPU-bound work across cores. The skill covers both concurrency approaches and mixed workloads that combine I/O and CPU-intensive tasks across Python 3.13+ projects.

Database optimization strategies include profiling query execution time, analyzing ORM usage patterns, and identifying N+1 query problems. This skill provides targeted patterns for improving query performance and reducing database round trips in production applications.

Py-spy samples running processes without code modification, making it ideal for profiling long-running or production services where cProfile overhead matters. cProfile requires code instrumentation but provides detailed function-level metrics; py-spy offers lower-overhead statistical sampling.