blob: bd67fe486abf7779f15c8881561299337e39b894 [file] [log] [blame]
Georg Brandl116aa622007-08-15 14:28:22 +00001.. _profile:
2
3********************
4The Python Profilers
5********************
6
Raymond Hettinger469271d2011-01-27 20:38:46 +00007**Source code:** :source:`Lib/profile.py` and :source:`Lib/pstats.py`
8
9--------------
Georg Brandl116aa622007-08-15 14:28:22 +000010
Georg Brandl116aa622007-08-15 14:28:22 +000011.. _profiler-introduction:
12
13Introduction to the profilers
14=============================
15
16.. index::
17 single: deterministic profiling
18 single: profiling, deterministic
19
Ezio Melotti075d87c2013-04-12 15:42:06 +030020:mod:`cProfile` and :mod:`profile` provide :dfn:`deterministic profiling` of
21Python programs. A :dfn:`profile` is a set of statistics that describes how
22often and for how long various parts of the program executed. These statistics
23can be formatted into reports via the :mod:`pstats` module.
Georg Brandl116aa622007-08-15 14:28:22 +000024
Ezio Melotti075d87c2013-04-12 15:42:06 +030025The Python standard library provides two different implementations of the same
26profiling interface:
Georg Brandl116aa622007-08-15 14:28:22 +000027
Georg Brandl4eb65972010-10-14 06:41:42 +0000281. :mod:`cProfile` is recommended for most users; it's a C extension with
29 reasonable overhead that makes it suitable for profiling long-running
30 programs. Based on :mod:`lsprof`, contributed by Brett Rosen and Ted
31 Czotter.
Georg Brandl116aa622007-08-15 14:28:22 +000032
Georg Brandl4eb65972010-10-14 06:41:42 +0000332. :mod:`profile`, a pure Python module whose interface is imitated by
Ezio Melotti075d87c2013-04-12 15:42:06 +030034 :mod:`cProfile`, but which adds significant overhead to profiled programs.
35 If you're trying to extend the profiler in some way, the task might be easier
Guido van Rossum6eb740b2015-09-10 12:12:01 -070036 with this module. Originally designed and written by Jim Roskind.
Georg Brandl4eb65972010-10-14 06:41:42 +000037
38.. note::
39
40 The profiler modules are designed to provide an execution profile for a given
41 program, not for benchmarking purposes (for that, there is :mod:`timeit` for
Ezio Melottib5ff3e42011-04-03 16:20:21 +030042 reasonably accurate results). This particularly applies to benchmarking
Georg Brandl4eb65972010-10-14 06:41:42 +000043 Python code against C code: the profilers introduce overhead for Python code,
44 but not for C-level functions, and so the C code would seem faster than any
45 Python one.
Georg Brandl116aa622007-08-15 14:28:22 +000046
47
48.. _profile-instant:
49
50Instant User's Manual
51=====================
52
53This section is provided for users that "don't want to read the manual." It
54provides a very brief overview, and allows a user to rapidly perform profiling
55on an existing application.
56
Ezio Melotti075d87c2013-04-12 15:42:06 +030057To profile a function that takes a single argument, you can do::
Georg Brandl116aa622007-08-15 14:28:22 +000058
59 import cProfile
Ezio Melotti075d87c2013-04-12 15:42:06 +030060 import re
61 cProfile.run('re.compile("foo|bar")')
Georg Brandl116aa622007-08-15 14:28:22 +000062
63(Use :mod:`profile` instead of :mod:`cProfile` if the latter is not available on
64your system.)
65
Ezio Melotti075d87c2013-04-12 15:42:06 +030066The above action would run :func:`re.compile` and print profile results like
67the following::
68
69 197 function calls (192 primitive calls) in 0.002 seconds
70
71 Ordered by: standard name
72
73 ncalls tottime percall cumtime percall filename:lineno(function)
74 1 0.000 0.000 0.001 0.001 <string>:1(<module>)
75 1 0.000 0.000 0.001 0.001 re.py:212(compile)
76 1 0.000 0.000 0.001 0.001 re.py:268(_compile)
77 1 0.000 0.000 0.000 0.000 sre_compile.py:172(_compile_charset)
78 1 0.000 0.000 0.000 0.000 sre_compile.py:201(_optimize_charset)
79 4 0.000 0.000 0.000 0.000 sre_compile.py:25(_identityfunction)
80 3/1 0.000 0.000 0.000 0.000 sre_compile.py:33(_compile)
81
82The first line indicates that 197 calls were monitored. Of those calls, 192
83were :dfn:`primitive`, meaning that the call was not induced via recursion. The
84next line: ``Ordered by: standard name``, indicates that the text string in the
85far right column was used to sort the output. The column headings include:
86
87ncalls
88 for the number of calls,
89
90tottime
91 for the total time spent in the given function (and excluding time made in
92 calls to sub-functions)
93
94percall
95 is the quotient of ``tottime`` divided by ``ncalls``
96
97cumtime
98 is the cumulative time spent in this and all subfunctions (from invocation
99 till exit). This figure is accurate *even* for recursive functions.
100
101percall
102 is the quotient of ``cumtime`` divided by primitive calls
103
104filename:lineno(function)
105 provides the respective data of each function
106
107When there are two numbers in the first column (for example ``3/1``), it means
108that the function recursed. The second value is the number of primitive calls
109and the former is the total number of calls. Note that when the function does
110not recurse, these two values are the same, and only the single figure is
111printed.
112
113Instead of printing the output at the end of the profile run, you can save the
114results to a file by specifying a filename to the :func:`run` function::
Georg Brandl116aa622007-08-15 14:28:22 +0000115
116 import cProfile
Ezio Melotti075d87c2013-04-12 15:42:06 +0300117 import re
118 cProfile.run('re.compile("foo|bar")', 'restats')
Georg Brandl116aa622007-08-15 14:28:22 +0000119
Ezio Melotti075d87c2013-04-12 15:42:06 +0300120The :class:`pstats.Stats` class reads profile results from a file and formats
121them in various ways.
122
123The file :mod:`cProfile` can also be invoked as a script to profile another
Georg Brandl116aa622007-08-15 14:28:22 +0000124script. For example::
125
Ezio Melotti075d87c2013-04-12 15:42:06 +0300126 python -m cProfile [-o output_file] [-s sort_order] myscript.py
Georg Brandl116aa622007-08-15 14:28:22 +0000127
Ezio Melotti075d87c2013-04-12 15:42:06 +0300128``-o`` writes the profile results to a file instead of to stdout
Georg Brandl116aa622007-08-15 14:28:22 +0000129
Ezio Melotti075d87c2013-04-12 15:42:06 +0300130``-s`` specifies one of the :func:`~pstats.Stats.sort_stats` sort values to sort
131the output by. This only applies when ``-o`` is not supplied.
Georg Brandl116aa622007-08-15 14:28:22 +0000132
Ezio Melotti075d87c2013-04-12 15:42:06 +0300133The :mod:`pstats` module's :class:`~pstats.Stats` class has a variety of methods
134for manipulating and printing the data saved into a profile results file::
Georg Brandl116aa622007-08-15 14:28:22 +0000135
136 import pstats
Ezio Melotti075d87c2013-04-12 15:42:06 +0300137 p = pstats.Stats('restats')
Georg Brandl116aa622007-08-15 14:28:22 +0000138 p.strip_dirs().sort_stats(-1).print_stats()
139
Ezio Melotti075d87c2013-04-12 15:42:06 +0300140The :meth:`~pstats.Stats.strip_dirs` method removed the extraneous path from all
141the module names. The :meth:`~pstats.Stats.sort_stats` method sorted all the
142entries according to the standard module/line/name string that is printed. The
143:meth:`~pstats.Stats.print_stats` method printed out all the statistics. You
144might try the following sort calls::
Georg Brandl116aa622007-08-15 14:28:22 +0000145
146 p.sort_stats('name')
147 p.print_stats()
148
149The first call will actually sort the list by function name, and the second call
150will print out the statistics. The following are some interesting calls to
151experiment with::
152
153 p.sort_stats('cumulative').print_stats(10)
154
155This sorts the profile by cumulative time in a function, and then only prints
156the ten most significant lines. If you want to understand what algorithms are
157taking time, the above line is what you would use.
158
159If you were looking to see what functions were looping a lot, and taking a lot
160of time, you would do::
161
162 p.sort_stats('time').print_stats(10)
163
164to sort according to time spent within each function, and then print the
165statistics for the top ten functions.
166
167You might also try::
168
169 p.sort_stats('file').print_stats('__init__')
170
171This will sort all the statistics by file name, and then print out statistics
172for only the class init methods (since they are spelled with ``__init__`` in
173them). As one final example, you could try::
174
Berker Peksag25587742015-06-05 14:48:29 +0300175 p.sort_stats('time', 'cumulative').print_stats(.5, 'init')
Georg Brandl116aa622007-08-15 14:28:22 +0000176
177This line sorts statistics with a primary key of time, and a secondary key of
178cumulative time, and then prints out some of the statistics. To be specific, the
179list is first culled down to 50% (re: ``.5``) of its original size, then only
180lines containing ``init`` are maintained, and that sub-sub-list is printed.
181
182If you wondered what functions called the above functions, you could now (``p``
183is still sorted according to the last criteria) do::
184
185 p.print_callers(.5, 'init')
186
187and you would get a list of callers for each of the listed functions.
188
189If you want more functionality, you're going to have to read the manual, or
190guess what the following functions do::
191
192 p.print_callees()
Ezio Melotti075d87c2013-04-12 15:42:06 +0300193 p.add('restats')
Georg Brandl116aa622007-08-15 14:28:22 +0000194
195Invoked as a script, the :mod:`pstats` module is a statistics browser for
196reading and examining profile dumps. It has a simple line-oriented interface
197(implemented using :mod:`cmd`) and interactive help.
198
Ezio Melotti075d87c2013-04-12 15:42:06 +0300199:mod:`profile` and :mod:`cProfile` Module Reference
200=======================================================
201
202.. module:: cProfile
203.. module:: profile
204 :synopsis: Python source profiler.
205
206Both the :mod:`profile` and :mod:`cProfile` modules provide the following
207functions:
208
209.. function:: run(command, filename=None, sort=-1)
210
211 This function takes a single argument that can be passed to the :func:`exec`
212 function, and an optional file name. In all cases this routine executes::
213
214 exec(command, __main__.__dict__, __main__.__dict__)
215
216 and gathers profiling statistics from the execution. If no file name is
217 present, then this function automatically creates a :class:`~pstats.Stats`
218 instance and prints a simple profiling report. If the sort value is specified
219 it is passed to this :class:`~pstats.Stats` instance to control how the
220 results are sorted.
221
222.. function:: runctx(command, globals, locals, filename=None)
223
224 This function is similar to :func:`run`, with added arguments to supply the
225 globals and locals dictionaries for the *command* string. This routine
226 executes::
227
228 exec(command, globals, locals)
229
230 and gathers profiling statistics as in the :func:`run` function above.
231
232.. class:: Profile(timer=None, timeunit=0.0, subcalls=True, builtins=True)
233
234 This class is normally only used if more precise control over profiling is
235 needed than what the :func:`cProfile.run` function provides.
236
237 A custom timer can be supplied for measuring how long code takes to run via
238 the *timer* argument. This must be a function that returns a single number
239 representing the current time. If the number is an integer, the *timeunit*
240 specifies a multiplier that specifies the duration of each unit of time. For
241 example, if the timer returns times measured in thousands of seconds, the
242 time unit would be ``.001``.
243
244 Directly using the :class:`Profile` class allows formatting profile results
245 without writing the profile data to a file::
246
247 import cProfile, pstats, io
248 pr = cProfile.Profile()
249 pr.enable()
Senthil Kumaran21101f72013-09-07 17:51:58 -0700250 # ... do something ...
Ezio Melotti075d87c2013-04-12 15:42:06 +0300251 pr.disable()
252 s = io.StringIO()
Senthil Kumaran21101f72013-09-07 17:51:58 -0700253 sortby = 'cumulative'
254 ps = pstats.Stats(pr, stream=s).sort_stats(sortby)
255 ps.print_stats()
256 print(s.getvalue())
Ezio Melotti075d87c2013-04-12 15:42:06 +0300257
258 .. method:: enable()
259
260 Start collecting profiling data.
261
262 .. method:: disable()
263
264 Stop collecting profiling data.
265
266 .. method:: create_stats()
267
268 Stop collecting profiling data and record the results internally
269 as the current profile.
270
271 .. method:: print_stats(sort=-1)
272
273 Create a :class:`~pstats.Stats` object based on the current
274 profile and print the results to stdout.
275
276 .. method:: dump_stats(filename)
277
278 Write the results of the current profile to *filename*.
279
280 .. method:: run(cmd)
281
282 Profile the cmd via :func:`exec`.
283
284 .. method:: runctx(cmd, globals, locals)
285
286 Profile the cmd via :func:`exec` with the specified global and
287 local environment.
288
289 .. method:: runcall(func, *args, **kwargs)
290
291 Profile ``func(*args, **kwargs)``
292
293.. _profile-stats:
294
295The :class:`Stats` Class
296========================
297
298Analysis of the profiler data is done using the :class:`~pstats.Stats` class.
299
300.. module:: pstats
301 :synopsis: Statistics object for use with the profiler.
302
303.. class:: Stats(*filenames or profile, stream=sys.stdout)
304
305 This class constructor creates an instance of a "statistics object" from a
306 *filename* (or list of filenames) or from a :class:`Profile` instance. Output
307 will be printed to the stream specified by *stream*.
308
309 The file selected by the above constructor must have been created by the
310 corresponding version of :mod:`profile` or :mod:`cProfile`. To be specific,
311 there is *no* file compatibility guaranteed with future versions of this
312 profiler, and there is no compatibility with files produced by other
313 profilers. If several files are provided, all the statistics for identical
314 functions will be coalesced, so that an overall view of several processes can
315 be considered in a single report. If additional files need to be combined
316 with data in an existing :class:`~pstats.Stats` object, the
317 :meth:`~pstats.Stats.add` method can be used.
318
319 Instead of reading the profile data from a file, a :class:`cProfile.Profile`
320 or :class:`profile.Profile` object can be used as the profile data source.
321
322 :class:`Stats` objects have the following methods:
323
324 .. method:: strip_dirs()
325
326 This method for the :class:`Stats` class removes all leading path
327 information from file names. It is very useful in reducing the size of
328 the printout to fit within (close to) 80 columns. This method modifies
329 the object, and the stripped information is lost. After performing a
330 strip operation, the object is considered to have its entries in a
331 "random" order, as it was just after object initialization and loading.
332 If :meth:`~pstats.Stats.strip_dirs` causes two function names to be
333 indistinguishable (they are on the same line of the same filename, and
334 have the same function name), then the statistics for these two entries
335 are accumulated into a single entry.
336
337
338 .. method:: add(*filenames)
339
340 This method of the :class:`Stats` class accumulates additional profiling
341 information into the current profiling object. Its arguments should refer
342 to filenames created by the corresponding version of :func:`profile.run`
343 or :func:`cProfile.run`. Statistics for identically named (re: file, line,
344 name) functions are automatically accumulated into single function
345 statistics.
346
347
348 .. method:: dump_stats(filename)
349
350 Save the data loaded into the :class:`Stats` object to a file named
351 *filename*. The file is created if it does not exist, and is overwritten
352 if it already exists. This is equivalent to the method of the same name
353 on the :class:`profile.Profile` and :class:`cProfile.Profile` classes.
354
355
356 .. method:: sort_stats(*keys)
357
358 This method modifies the :class:`Stats` object by sorting it according to
359 the supplied criteria. The argument is typically a string identifying the
360 basis of a sort (example: ``'time'`` or ``'name'``).
361
362 When more than one key is provided, then additional keys are used as
363 secondary criteria when there is equality in all keys selected before
364 them. For example, ``sort_stats('name', 'file')`` will sort all the
365 entries according to their function name, and resolve all ties (identical
366 function names) by sorting by file name.
367
368 Abbreviations can be used for any key names, as long as the abbreviation
369 is unambiguous. The following are the keys currently defined:
370
371 +------------------+----------------------+
372 | Valid Arg | Meaning |
373 +==================+======================+
374 | ``'calls'`` | call count |
375 +------------------+----------------------+
376 | ``'cumulative'`` | cumulative time |
377 +------------------+----------------------+
378 | ``'cumtime'`` | cumulative time |
379 +------------------+----------------------+
380 | ``'file'`` | file name |
381 +------------------+----------------------+
382 | ``'filename'`` | file name |
383 +------------------+----------------------+
384 | ``'module'`` | file name |
385 +------------------+----------------------+
386 | ``'ncalls'`` | call count |
387 +------------------+----------------------+
388 | ``'pcalls'`` | primitive call count |
389 +------------------+----------------------+
390 | ``'line'`` | line number |
391 +------------------+----------------------+
392 | ``'name'`` | function name |
393 +------------------+----------------------+
394 | ``'nfl'`` | name/file/line |
395 +------------------+----------------------+
396 | ``'stdname'`` | standard name |
397 +------------------+----------------------+
398 | ``'time'`` | internal time |
399 +------------------+----------------------+
400 | ``'tottime'`` | internal time |
401 +------------------+----------------------+
402
403 Note that all sorts on statistics are in descending order (placing most
404 time consuming items first), where as name, file, and line number searches
405 are in ascending order (alphabetical). The subtle distinction between
406 ``'nfl'`` and ``'stdname'`` is that the standard name is a sort of the
407 name as printed, which means that the embedded line numbers get compared
408 in an odd way. For example, lines 3, 20, and 40 would (if the file names
409 were the same) appear in the string order 20, 3 and 40. In contrast,
410 ``'nfl'`` does a numeric compare of the line numbers. In fact,
411 ``sort_stats('nfl')`` is the same as ``sort_stats('name', 'file',
412 'line')``.
413
414 For backward-compatibility reasons, the numeric arguments ``-1``, ``0``,
415 ``1``, and ``2`` are permitted. They are interpreted as ``'stdname'``,
416 ``'calls'``, ``'time'``, and ``'cumulative'`` respectively. If this old
417 style format (numeric) is used, only one sort key (the numeric key) will
418 be used, and additional arguments will be silently ignored.
419
420 .. For compatibility with the old profiler.
421
422
423 .. method:: reverse_order()
424
425 This method for the :class:`Stats` class reverses the ordering of the
426 basic list within the object. Note that by default ascending vs
427 descending order is properly selected based on the sort key of choice.
428
429 .. This method is provided primarily for compatibility with the old
430 profiler.
431
432
433 .. method:: print_stats(*restrictions)
434
435 This method for the :class:`Stats` class prints out a report as described
436 in the :func:`profile.run` definition.
437
438 The order of the printing is based on the last
439 :meth:`~pstats.Stats.sort_stats` operation done on the object (subject to
440 caveats in :meth:`~pstats.Stats.add` and
441 :meth:`~pstats.Stats.strip_dirs`).
442
443 The arguments provided (if any) can be used to limit the list down to the
444 significant entries. Initially, the list is taken to be the complete set
445 of profiled functions. Each restriction is either an integer (to select a
446 count of lines), or a decimal fraction between 0.0 and 1.0 inclusive (to
Berker Peksagb067a5e2017-02-22 04:55:33 +0300447 select a percentage of lines), or a string that will interpreted as a
448 regular expression (to pattern match the standard name that is printed).
449 If several restrictions are provided, then they are applied sequentially.
450 For example::
Ezio Melotti075d87c2013-04-12 15:42:06 +0300451
452 print_stats(.1, 'foo:')
453
454 would first limit the printing to first 10% of list, and then only print
455 functions that were part of filename :file:`.\*foo:`. In contrast, the
456 command::
457
458 print_stats('foo:', .1)
459
460 would limit the list to all functions having file names :file:`.\*foo:`,
461 and then proceed to only print the first 10% of them.
462
463
464 .. method:: print_callers(*restrictions)
465
466 This method for the :class:`Stats` class prints a list of all functions
467 that called each function in the profiled database. The ordering is
468 identical to that provided by :meth:`~pstats.Stats.print_stats`, and the
469 definition of the restricting argument is also identical. Each caller is
470 reported on its own line. The format differs slightly depending on the
471 profiler that produced the stats:
472
473 * With :mod:`profile`, a number is shown in parentheses after each caller
474 to show how many times this specific call was made. For convenience, a
475 second non-parenthesized number repeats the cumulative time spent in the
476 function at the right.
477
478 * With :mod:`cProfile`, each caller is preceded by three numbers: the
479 number of times this specific call was made, and the total and
480 cumulative times spent in the current function while it was invoked by
481 this specific caller.
482
483
484 .. method:: print_callees(*restrictions)
485
486 This method for the :class:`Stats` class prints a list of all function
487 that were called by the indicated function. Aside from this reversal of
488 direction of calls (re: called vs was called by), the arguments and
489 ordering are identical to the :meth:`~pstats.Stats.print_callers` method.
490
Georg Brandl116aa622007-08-15 14:28:22 +0000491
492.. _deterministic-profiling:
493
494What Is Deterministic Profiling?
495================================
496
497:dfn:`Deterministic profiling` is meant to reflect the fact that all *function
498call*, *function return*, and *exception* events are monitored, and precise
499timings are made for the intervals between these events (during which time the
500user's code is executing). In contrast, :dfn:`statistical profiling` (which is
501not done by this module) randomly samples the effective instruction pointer, and
502deduces where time is being spent. The latter technique traditionally involves
503less overhead (as the code does not need to be instrumented), but provides only
504relative indications of where time is being spent.
505
506In Python, since there is an interpreter active during execution, the presence
507of instrumented code is not required to do deterministic profiling. Python
508automatically provides a :dfn:`hook` (optional callback) for each event. In
509addition, the interpreted nature of Python tends to add so much overhead to
510execution, that deterministic profiling tends to only add small processing
511overhead in typical applications. The result is that deterministic profiling is
512not that expensive, yet provides extensive run time statistics about the
513execution of a Python program.
514
515Call count statistics can be used to identify bugs in code (surprising counts),
516and to identify possible inline-expansion points (high call counts). Internal
517time statistics can be used to identify "hot loops" that should be carefully
518optimized. Cumulative time statistics should be used to identify high level
519errors in the selection of algorithms. Note that the unusual handling of
520cumulative times in this profiler allows statistics for recursive
521implementations of algorithms to be directly compared to iterative
522implementations.
523
524
Ezio Melotti075d87c2013-04-12 15:42:06 +0300525.. _profile-limitations:
Georg Brandl116aa622007-08-15 14:28:22 +0000526
527Limitations
528===========
529
530One limitation has to do with accuracy of timing information. There is a
531fundamental problem with deterministic profilers involving accuracy. The most
532obvious restriction is that the underlying "clock" is only ticking at a rate
533(typically) of about .001 seconds. Hence no measurements will be more accurate
534than the underlying clock. If enough measurements are taken, then the "error"
535will tend to average out. Unfortunately, removing this first error induces a
536second source of error.
537
538The second problem is that it "takes a while" from when an event is dispatched
539until the profiler's call to get the time actually *gets* the state of the
540clock. Similarly, there is a certain lag when exiting the profiler event
541handler from the time that the clock's value was obtained (and then squirreled
542away), until the user's code is once again executing. As a result, functions
543that are called many times, or call many functions, will typically accumulate
544this error. The error that accumulates in this fashion is typically less than
545the accuracy of the clock (less than one clock tick), but it *can* accumulate
546and become very significant.
547
548The problem is more important with :mod:`profile` than with the lower-overhead
549:mod:`cProfile`. For this reason, :mod:`profile` provides a means of
550calibrating itself for a given platform so that this error can be
551probabilistically (on the average) removed. After the profiler is calibrated, it
552will be more accurate (in a least square sense), but it will sometimes produce
553negative numbers (when call counts are exceptionally low, and the gods of
554probability work against you :-). ) Do *not* be alarmed by negative numbers in
555the profile. They should *only* appear if you have calibrated your profiler,
556and the results are actually better than without calibration.
557
558
559.. _profile-calibration:
560
561Calibration
562===========
563
564The profiler of the :mod:`profile` module subtracts a constant from each event
565handling time to compensate for the overhead of calling the time function, and
566socking away the results. By default, the constant is 0. The following
567procedure can be used to obtain a better constant for a given platform (see
Ezio Melotti075d87c2013-04-12 15:42:06 +0300568:ref:`profile-limitations`). ::
Georg Brandl116aa622007-08-15 14:28:22 +0000569
570 import profile
571 pr = profile.Profile()
572 for i in range(5):
Georg Brandl6911e3c2007-09-04 07:15:32 +0000573 print(pr.calibrate(10000))
Georg Brandl116aa622007-08-15 14:28:22 +0000574
575The method executes the number of Python calls given by the argument, directly
576and again under the profiler, measuring the time for both. It then computes the
577hidden overhead per profiler event, and returns that as a float. For example,
Ezio Melotti075d87c2013-04-12 15:42:06 +0300578on a 1.8Ghz Intel Core i5 running Mac OS X, and using Python's time.clock() as
579the timer, the magical number is about 4.04e-6.
Georg Brandl116aa622007-08-15 14:28:22 +0000580
581The object of this exercise is to get a fairly consistent result. If your
582computer is *very* fast, or your timer function has poor resolution, you might
583have to pass 100000, or even 1000000, to get consistent results.
584
Georg Brandle6bcc912008-05-12 18:05:20 +0000585When you have a consistent answer, there are three ways you can use it::
Georg Brandl116aa622007-08-15 14:28:22 +0000586
587 import profile
588
589 # 1. Apply computed bias to all Profile instances created hereafter.
590 profile.Profile.bias = your_computed_bias
591
592 # 2. Apply computed bias to a specific Profile instance.
593 pr = profile.Profile()
594 pr.bias = your_computed_bias
595
596 # 3. Specify computed bias in instance constructor.
597 pr = profile.Profile(bias=your_computed_bias)
598
599If you have a choice, you are better off choosing a smaller constant, and then
600your results will "less often" show up as negative in profile statistics.
601
Ezio Melotti075d87c2013-04-12 15:42:06 +0300602.. _profile-timers:
Georg Brandl116aa622007-08-15 14:28:22 +0000603
Georg Brandlc2b17b22013-10-06 09:17:43 +0200604Using a custom timer
605====================
Georg Brandl116aa622007-08-15 14:28:22 +0000606
Ezio Melotti075d87c2013-04-12 15:42:06 +0300607If you want to change how current time is determined (for example, to force use
608of wall-clock time or elapsed process time), pass the timing function you want
609to the :class:`Profile` class constructor::
Georg Brandl116aa622007-08-15 14:28:22 +0000610
Ezio Melotti075d87c2013-04-12 15:42:06 +0300611 pr = profile.Profile(your_time_func)
Georg Brandl116aa622007-08-15 14:28:22 +0000612
Ezio Melotti075d87c2013-04-12 15:42:06 +0300613The resulting profiler will then call ``your_time_func``. Depending on whether
614you are using :class:`profile.Profile` or :class:`cProfile.Profile`,
615``your_time_func``'s return value will be interpreted differently:
Georg Brandl116aa622007-08-15 14:28:22 +0000616
617:class:`profile.Profile`
Ezio Melotti075d87c2013-04-12 15:42:06 +0300618 ``your_time_func`` should return a single number, or a list of numbers whose
619 sum is the current time (like what :func:`os.times` returns). If the
620 function returns a single time number, or the list of returned numbers has
621 length 2, then you will get an especially fast version of the dispatch
622 routine.
Georg Brandl116aa622007-08-15 14:28:22 +0000623
Ezio Melotti075d87c2013-04-12 15:42:06 +0300624 Be warned that you should calibrate the profiler class for the timer function
625 that you choose (see :ref:`profile-calibration`). For most machines, a timer
626 that returns a lone integer value will provide the best results in terms of
627 low overhead during profiling. (:func:`os.times` is *pretty* bad, as it
628 returns a tuple of floating point values). If you want to substitute a
629 better timer in the cleanest fashion, derive a class and hardwire a
630 replacement dispatch method that best handles your timer call, along with the
631 appropriate calibration constant.
Georg Brandl116aa622007-08-15 14:28:22 +0000632
633:class:`cProfile.Profile`
Ezio Melotti075d87c2013-04-12 15:42:06 +0300634 ``your_time_func`` should return a single number. If it returns integers,
635 you can also invoke the class constructor with a second argument specifying
636 the real duration of one unit of time. For example, if
637 ``your_integer_time_func`` returns times measured in thousands of seconds,
638 you would construct the :class:`Profile` instance as follows::
Georg Brandl116aa622007-08-15 14:28:22 +0000639
Ezio Melotti075d87c2013-04-12 15:42:06 +0300640 pr = cProfile.Profile(your_integer_time_func, 0.001)
Georg Brandl116aa622007-08-15 14:28:22 +0000641
Serhiy Storchakab19542d2015-03-14 21:32:57 +0200642 As the :class:`cProfile.Profile` class cannot be calibrated, custom timer
Ezio Melotti075d87c2013-04-12 15:42:06 +0300643 functions should be used with care and should be as fast as possible. For
644 the best results with a custom timer, it might be necessary to hard-code it
645 in the C source of the internal :mod:`_lsprof` module.
646
647Python 3.3 adds several new functions in :mod:`time` that can be used to make
648precise measurements of process or wall-clock time. For example, see
649:func:`time.perf_counter`.