suggest-optimizations

Performance bottlenecks are functions or code paths consuming disproportionate time or memory. Use profiling with cProfile to measure execution data, then analyze the results to pinpoint which functions deserve optimization effort and deliver the highest performance gains.

Common optimization techniques include algorithmic improvements, data structure changes, caching, and SIMD vectorization. Evaluate each candidate by expected performance impact versus implementation difficulty, then apply the highest-impact, lowest-effort optimizations first.

Yes. Profiling reveals which data structures and algorithms consume the most resources in your actual workload. Analyze hot paths from profile output to identify opportunities for more efficient algorithms or better-suited data structures with measurable performance improvement.

Identify performance-critical paths using profiling, then assess whether vectorization applies—typically for numeric operations on large datasets. Plan SIMD adoption by measuring baseline performance, selecting appropriate vectorization approaches, and validating improvements against profiling benchmarks.

Profile your code with cProfile to gather execution data, identify the top time and memory consumers, analyze their implementation for optimization opportunities, propose specific techniques with expected impact and difficulty estimates, then implement and re-profile to validate improvements.