未翻訳のページ
このページはまだ日本語に翻訳されていないため、英語の原文を表示しています。 翻訳に協力する
cProfile Module¶
The cProfile module provides deterministic profiling of Python programs, measuring function call frequency and execution time.
Complexity Reference¶
| Operation | Time | Space | Notes |
|---|---|---|---|
cProfile.run() |
O(n) | O(n) | n = function calls |
cProfile.runctx() |
O(n) | O(n) | n = function calls; with globals/locals |
cProfile.main() |
O(n) | O(n) | CLI entrypoint |
Profile() |
O(1) | O(1) | Profiler instance |
| Profile collection | O(1) | O(1) | Per call overhead |
| Report generation | O(n log n) | O(n) | n = unique functions |
Profiling Functions¶
Basic Profiling¶
import cProfile
def fibonacci(n):
if n <= 1:
return n
return fibonacci(n-1) + fibonacci(n-2)
# Profile function - O(n) where n = calls
cProfile.run('fibonacci(10)')
# Output shows: ncalls, tottime, cumtime, filename:lineno(function)
Analyzing Profile Data¶
import cProfile
import pstats
# Profile with save - O(n)
cProfile.run('fibonacci(10)', 'profile_stats')
# Load and analyze - O(n log n)
stats = pstats.Stats('profile_stats')
stats.strip_dirs()
stats.sort_stats('cumulative')
stats.print_stats(10) # Top 10