Vladimir Sementsov-Ogievskiy | 7cc8e0a | 2020-02-28 10:19:11 +0300 | [diff] [blame] | 1 | #!/usr/bin/env python |
| 2 | # |
| 3 | # Simple benchmarking framework |
| 4 | # |
| 5 | # Copyright (c) 2019 Virtuozzo International GmbH. |
| 6 | # |
| 7 | # This program is free software; you can redistribute it and/or modify |
| 8 | # it under the terms of the GNU General Public License as published by |
| 9 | # the Free Software Foundation; either version 2 of the License, or |
| 10 | # (at your option) any later version. |
| 11 | # |
| 12 | # This program is distributed in the hope that it will be useful, |
| 13 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 15 | # GNU General Public License for more details. |
| 16 | # |
| 17 | # You should have received a copy of the GNU General Public License |
| 18 | # along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 19 | # |
| 20 | |
| 21 | |
| 22 | def bench_one(test_func, test_env, test_case, count=5, initial_run=True): |
| 23 | """Benchmark one test-case |
| 24 | |
| 25 | test_func -- benchmarking function with prototype |
| 26 | test_func(env, case), which takes test_env and test_case |
| 27 | arguments and returns {'seconds': int} (which is benchmark |
| 28 | result) on success and {'error': str} on error. Returned |
| 29 | dict may contain any other additional fields. |
| 30 | test_env -- test environment - opaque first argument for test_func |
| 31 | test_case -- test case - opaque second argument for test_func |
| 32 | count -- how many times to call test_func, to calculate average |
| 33 | initial_run -- do initial run of test_func, which don't get into result |
| 34 | |
| 35 | Returns dict with the following fields: |
| 36 | 'runs': list of test_func results |
| 37 | 'average': average seconds per run (exists only if at least one run |
| 38 | succeeded) |
| 39 | 'delta': maximum delta between test_func result and the average |
| 40 | (exists only if at least one run succeeded) |
| 41 | 'n-failed': number of failed runs (exists only if at least one run |
| 42 | failed) |
| 43 | """ |
| 44 | if initial_run: |
| 45 | print(' #initial run:') |
| 46 | print(' ', test_func(test_env, test_case)) |
| 47 | |
| 48 | runs = [] |
| 49 | for i in range(count): |
| 50 | print(' #run {}'.format(i+1)) |
| 51 | res = test_func(test_env, test_case) |
| 52 | print(' ', res) |
| 53 | runs.append(res) |
| 54 | |
| 55 | result = {'runs': runs} |
| 56 | |
| 57 | successed = [r for r in runs if ('seconds' in r)] |
| 58 | if successed: |
| 59 | avg = sum(r['seconds'] for r in successed) / len(successed) |
| 60 | result['average'] = avg |
| 61 | result['delta'] = max(abs(r['seconds'] - avg) for r in successed) |
| 62 | |
| 63 | if len(successed) < count: |
| 64 | result['n-failed'] = count - len(successed) |
| 65 | |
| 66 | return result |
| 67 | |
| 68 | |
| 69 | def ascii_one(result): |
| 70 | """Return ASCII representation of bench_one() returned dict.""" |
| 71 | if 'average' in result: |
| 72 | s = '{:.2f} +- {:.2f}'.format(result['average'], result['delta']) |
| 73 | if 'n-failed' in result: |
| 74 | s += '\n({} failed)'.format(result['n-failed']) |
| 75 | return s |
| 76 | else: |
| 77 | return 'FAILED' |
| 78 | |
| 79 | |
| 80 | def bench(test_func, test_envs, test_cases, *args, **vargs): |
| 81 | """Fill benchmark table |
| 82 | |
| 83 | test_func -- benchmarking function, see bench_one for description |
| 84 | test_envs -- list of test environments, see bench_one |
| 85 | test_cases -- list of test cases, see bench_one |
| 86 | args, vargs -- additional arguments for bench_one |
| 87 | |
| 88 | Returns dict with the following fields: |
| 89 | 'envs': test_envs |
| 90 | 'cases': test_cases |
| 91 | 'tab': filled 2D array, where cell [i][j] is bench_one result for |
| 92 | test_cases[i] for test_envs[j] (i.e., rows are test cases and |
| 93 | columns are test environments) |
| 94 | """ |
| 95 | tab = {} |
| 96 | results = { |
| 97 | 'envs': test_envs, |
| 98 | 'cases': test_cases, |
| 99 | 'tab': tab |
| 100 | } |
| 101 | n = 1 |
| 102 | n_tests = len(test_envs) * len(test_cases) |
| 103 | for env in test_envs: |
| 104 | for case in test_cases: |
| 105 | print('Testing {}/{}: {} :: {}'.format(n, n_tests, |
| 106 | env['id'], case['id'])) |
| 107 | if case['id'] not in tab: |
| 108 | tab[case['id']] = {} |
| 109 | tab[case['id']][env['id']] = bench_one(test_func, env, case, |
| 110 | *args, **vargs) |
| 111 | n += 1 |
| 112 | |
| 113 | print('Done') |
| 114 | return results |
| 115 | |
| 116 | |
| 117 | def ascii(results): |
| 118 | """Return ASCII representation of bench() returned dict.""" |
| 119 | from tabulate import tabulate |
| 120 | |
| 121 | tab = [[""] + [c['id'] for c in results['envs']]] |
| 122 | for case in results['cases']: |
| 123 | row = [case['id']] |
| 124 | for env in results['envs']: |
| 125 | row.append(ascii_one(results['tab'][case['id']][env['id']])) |
| 126 | tab.append(row) |
| 127 | |
| 128 | return tabulate(tab) |