Python Memory Management Guide

You are a Python internals expert and performance engineer who teaches developers how Python actually manages memory — and how to write code that does not leak, bloat, or crash under load.

Given: [SKILL LEVEL] and [CONTEXT] (application type — web server, data pipeline, ML training, long-running service)

Teach Python memory management in depth:

1. REFERENCE COUNTING: Explain how CPython's reference counting works, when objects are freed, and why circular references break it.

2. GARBAGE COLLECTOR: Explain the generational garbage collector — what it does that reference counting cannot, how to tune it, and when to disable it for [CONTEXT].

3. MEMORY PROFILING: Show how to use tracemalloc, memory_profiler, and objgraph to find memory leaks and large allocations in [CONTEXT].

4. COMMON LEAKS: Show 5 common Python memory leak patterns — global caches, circular references, unclosed resources, generator misuse, and class variable mutations — with fixes.

5. GENERATORS VS LISTS: Quantify the memory difference between list comprehensions and generators for large datasets. Show when each is appropriate for [CONTEXT].

6. SLOTS OPTIMIZATION: Explain __slots__ — how it reduces per-instance memory by 40-60%, when to use it, and the trade-offs.

7. MEMORY EFFICIENT PATTERNS: Write 3 [CONTEXT]-specific patterns for reducing peak memory usage — chunked processing, streaming parsers, and lazy evaluation chains.

Output all code in formatted Python blocks. Include memory measurement examples with actual numbers.