mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-09-26 18:29:57 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 10

#15979: improve timeit documentation.

Browse source
This commit is contained in:
Ezio Melotti 2012-10-02 05:34:38 +03:00
parent 71d74b0c4e
commit 31a9e83d7d
2 changed files with 180 additions and 114 deletions
Download patch file Download diff file Expand all files Collapse all files

292
Doc/library/timeit.rst
View file

@ -16,122 +16,163 @@
-------------- --------------
This module provides a simple way to time small bits of Python code. It has both This module provides a simple way to time small bits of Python code. It has both
command line as well as callable interfaces. It avoids a number of common traps a :ref:`command-line-interface` as well as a :ref:`callable <python-interface>`
for measuring execution times. See also Tim Peters' introduction to the one. It avoids a number of common traps for measuring execution times.
"Algorithms" chapter in the Python Cookbook, published by O'Reilly. See also Tim Peters' introduction to the "Algorithms" chapter in the *Python
Cookbook*, published by O'Reilly.
The module defines the following public class:
Basic Examples
--------------
The following example shows how the :ref:`command-line-interface`
can be used to compare three different expressions:
.. code-block:: sh
$ python -m timeit '"-".join(str(n) for n in range(100))'
10000 loops, best of 3: 40.3 usec per loop
$ python -m timeit '"-".join([str(n) for n in range(100)])'
10000 loops, best of 3: 33.4 usec per loop
$ python -m timeit '"-".join(map(str, range(100)))'
10000 loops, best of 3: 25.2 usec per loop
This can be achieved from the :ref:`python-interface` with::
>>> import timeit
>>> timeit.timeit('"-".join(str(n) for n in range(100))', number=10000)
0.8187260627746582
>>> timeit.timeit('"-".join([str(n) for n in range(100)])', number=10000)
0.7288308143615723
>>> timeit.timeit('"-".join(map(str, range(100)))', number=10000)
0.5858950614929199
Note however that :mod:`timeit` will automatically determine the number of
repetitions only when the command-line interface is used. In the
:ref:`timeit-examples` section you can find more advanced examples.
.. _python-interface:
Python Interface
----------------
The module defines three convenience functions and a public class:
.. function:: timeit(stmt='pass', setup='pass', timer=<default timer>, number=1000000)
Create a :class:`Timer` instance with the given statement, *setup* code and
*timer* function and run its :meth:`.timeit` method with *number* executions.
.. versionadded:: 2.6
.. function:: repeat(stmt='pass', setup='pass', timer=<default timer>, repeat=3, number=1000000)
Create a :class:`Timer` instance with the given statement, *setup* code and
*timer* function and run its :meth:`.repeat` method with the given *repeat*
count and *number* executions.
.. versionadded:: 2.6
.. function:: default_timer()
Define a default timer, in a platform-specific manner. On Windows,
:func:`time.clock` has microsecond granularity, but :func:`time.time`'s
granularity is 1/60th of a second. On Unix, :func:`time.clock` has 1/100th of
a second granularity, and :func:`time.time` is much more precise. On either
platform, :func:`default_timer` measures wall clock time, not the CPU
time. This means that other processes running on the same computer may
interfere with the timing.
.. class:: Timer(stmt='pass', setup='pass', timer=<timer function>) .. class:: Timer(stmt='pass', setup='pass', timer=<timer function>)
Class for timing execution speed of small code snippets. Class for timing execution speed of small code snippets.
The constructor takes a statement to be timed, an additional statement used for The constructor takes a statement to be timed, an additional statement used
setup, and a timer function. Both statements default to ``'pass'``; the timer for setup, and a timer function. Both statements default to ``'pass'``;
function is platform-dependent (see the module doc string). *stmt* and *setup* the timer function is platform-dependent (see the module doc string).
may also contain multiple statements separated by ``;`` or newlines, as long as *stmt* and *setup* may also contain multiple statements separated by ``;``
they don't contain multi-line string literals. or newlines, as long as they don't contain multi-line string literals.
To measure the execution time of the first statement, use the :meth:`Timer.timeit` To measure the execution time of the first statement, use the :meth:`.timeit`
method. The :meth:`repeat` method is a convenience to call :meth:`.timeit` method. The :meth:`.repeat` method is a convenience to call :meth:`.timeit`
multiple times and return a list of results. multiple times and return a list of results.
.. versionchanged:: 2.6 .. versionchanged:: 2.6
The *stmt* and *setup* parameters can now also take objects that are callable The *stmt* and *setup* parameters can now also take objects that are
without arguments. This will embed calls to them in a timer function that will callable without arguments. This will embed calls to them in a timer
then be executed by :meth:`.timeit`. Note that the timing overhead is a function that will then be executed by :meth:`.timeit`. Note that the
little larger in this case because of the extra function calls. timing overhead is a little larger in this case because of the extra
function calls.
.. method:: Timer.print_exc(file=None) .. method:: Timer.timeit(number=1000000)
Helper to print a traceback from the timed code. Time *number* executions of the main statement. This executes the setup
statement once, and then returns the time it takes to execute the main
statement a number of times, measured in seconds as a float.
The argument is the number of times through the loop, defaulting to one
million. The main statement, the setup statement and the timer function
to be used are passed to the constructor.
Typical use:: .. note::
t = Timer(...) # outside the try/except By default, :meth:`.timeit` temporarily turns off :term:`garbage
try: collection` during the timing. The advantage of this approach is that
t.timeit(...) # or t.repeat(...) it makes independent timings more comparable. This disadvantage is
except: that GC may be an important component of the performance of the
t.print_exc() function being measured. If so, GC can be re-enabled as the first
statement in the *setup* string. For example::
The advantage over the standard traceback is that source lines in the compiled timeit.Timer('for i in xrange(10): oct(i)', 'gc.enable()').timeit()
template will be displayed. The optional *file* argument directs where the
traceback is sent; it defaults to ``sys.stderr``.
.. method:: Timer.repeat(repeat=3, number=1000000) .. method:: Timer.repeat(repeat=3, number=1000000)
Call :meth:`.timeit` a few times. Call :meth:`.timeit` a few times.
This is a convenience function that calls the :meth:`.timeit` repeatedly, This is a convenience function that calls the :meth:`.timeit` repeatedly,
returning a list of results. The first argument specifies how many times to returning a list of results. The first argument specifies how many times
call :meth:`.timeit`. The second argument specifies the *number* argument for to call :meth:`.timeit`. The second argument specifies the *number*
:meth:`.timeit`. argument for :meth:`.timeit`.
.. note:: .. note::
It's tempting to calculate mean and standard deviation from the result vector It's tempting to calculate mean and standard deviation from the result
and report these. However, this is not very useful. In a typical case, the vector and report these. However, this is not very useful.
lowest value gives a lower bound for how fast your machine can run the given In a typical case, the lowest value gives a lower bound for how fast
code snippet; higher values in the result vector are typically not caused by your machine can run the given code snippet; higher values in the
variability in Python's speed, but by other processes interfering with your result vector are typically not caused by variability in Python's
timing accuracy. So the :func:`min` of the result is probably the only number speed, but by other processes interfering with your timing accuracy.
you should be interested in. After that, you should look at the entire vector So the :func:`min` of the result is probably the only number you
and apply common sense rather than statistics. should be interested in. After that, you should look at the entire
vector and apply common sense rather than statistics.
.. method:: Timer.timeit(number=1000000) .. method:: Timer.print_exc(file=None)
Time *number* executions of the main statement. This executes the setup Helper to print a traceback from the timed code.
statement once, and then returns the time it takes to execute the main statement
a number of times, measured in seconds as a float. The argument is the number
of times through the loop, defaulting to one million. The main statement, the
setup statement and the timer function to be used are passed to the constructor.
.. note:: Typical use::
By default, :meth:`.timeit` temporarily turns off :term:`garbage collection` t = Timer(...) # outside the try/except
during the timing. The advantage of this approach is that it makes try:
independent timings more comparable. This disadvantage is that GC may be t.timeit(...) # or t.repeat(...)
an important component of the performance of the function being measured. except:
If so, GC can be re-enabled as the first statement in the *setup* string. t.print_exc()
For example::
timeit.Timer('for i in xrange(10): oct(i)', 'gc.enable()').timeit() The advantage over the standard traceback is that source lines in the
compiled template will be displayed. The optional *file* argument directs
The module also defines three convenience functions: where the traceback is sent; it defaults to :data:`sys.stderr`.
.. function:: default_timer() .. _command-line-interface:
Define a default timer, in a platform specific manner. On Windows, Command-Line Interface
:func:`time.clock` has microsecond granularity but :func:`time.time`'s
granularity is 1/60th of a second; on Unix, :func:`time.clock` has 1/100th of
a second granularity and :func:`time.time` is much more precise. On either
platform, :func:`default_timer` measures wall clock time, not the CPU
time. This means that other processes running on the same computer may
interfere with the timing.
.. function:: repeat(stmt, setup='pass', timer=default_timer, repeat=3 , number=1000000)
Create a :class:`Timer` instance with the given statement, setup code and timer
function and run its :meth:`repeat` method with the given repeat count and
*number* executions.
.. versionadded:: 2.6
.. function:: timeit(stmt, setup='pass', timer=default_timer, number=1000000)
Create a :class:`Timer` instance with the given statement, setup code and timer
function and run its :meth:`.timeit` method with *number* executions.
.. versionadded:: 2.6
Command Line Interface
---------------------- ----------------------
When called as a program from the command line, the following form is used:: When called as a program from the command line, the following form is used::
@ -189,25 +230,55 @@ Unix, you can use :func:`time.clock` to measure CPU time.
There is a certain baseline overhead associated with executing a pass statement. There is a certain baseline overhead associated with executing a pass statement.
The code here doesn't try to hide it, but you should be aware of it. The The code here doesn't try to hide it, but you should be aware of it. The
baseline overhead can be measured by invoking the program without arguments. baseline overhead can be measured by invoking the program without arguments, and
it might differ between Python versions. Also, to fairly compare older Python
versions to Python 2.3, you may want to use Python's :option:`-O` option for
the older versions to avoid timing ``SET_LINENO`` instructions.
The baseline overhead differs between Python versions! Also, to fairly compare
older Python versions to Python 2.3, you may want to use Python's :option:`-O`
option for the older versions to avoid timing ``SET_LINENO`` instructions.
.. _timeit-examples:
Examples Examples
-------- --------
Here are two example sessions (one using the command line, one using the module It is possible to provide a setup statement that is executed only once at the beginning:
interface) that compare the cost of using :func:`hasattr` vs.
:keyword:`try`/:keyword:`except` to test for missing and present object .. code-block:: sh
attributes. ::
$ python -m timeit -s 'text = "sample string"; char = "g"' 'char in text'
10000000 loops, best of 3: 0.0877 usec per loop
$ python -m timeit -s 'text = "sample string"; char = "g"' 'text.find(char)'
1000000 loops, best of 3: 0.342 usec per loop
::
>>> import timeit
>>> timeit.timeit('char in text', setup='text = "sample string"; char = "g"')
0.41440500499993504
>>> timeit.timeit('text.find(char)', setup='text = "sample string"; char = "g"')
1.7246671520006203
The same can be done using the :class:`Timer` class and its methods::
>>> import timeit
>>> t = timeit.Timer('char in text', setup='text = "sample string"; char = "g"')
>>> t.timeit()
0.3955516149999312
>>> t.repeat()
[0.40193588800002544, 0.3960157959998014, 0.39594301399984033]
The following examples show how to time expressions that contain multiple lines.
Here we compare the cost of using :func:`hasattr` vs. :keyword:`try`/:keyword:`except`
to test for missing and present object attributes:
.. code-block:: sh
$ python -m timeit 'try:' ' str.__nonzero__' 'except AttributeError:' ' pass' $ python -m timeit 'try:' ' str.__nonzero__' 'except AttributeError:' ' pass'
100000 loops, best of 3: 15.7 usec per loop 100000 loops, best of 3: 15.7 usec per loop
$ python -m timeit 'if hasattr(str, "__nonzero__"): pass' $ python -m timeit 'if hasattr(str, "__nonzero__"): pass'
100000 loops, best of 3: 4.26 usec per loop 100000 loops, best of 3: 4.26 usec per loop
$ python -m timeit 'try:' ' int.__nonzero__' 'except AttributeError:' ' pass' $ python -m timeit 'try:' ' int.__nonzero__' 'except AttributeError:' ' pass'
1000000 loops, best of 3: 1.43 usec per loop 1000000 loops, best of 3: 1.43 usec per loop
$ python -m timeit 'if hasattr(int, "__nonzero__"): pass' $ python -m timeit 'if hasattr(int, "__nonzero__"): pass'
@ -216,36 +287,31 @@ attributes. ::
:: ::
>>> import timeit >>> import timeit
>>> # attribute is missing
>>> s = """\ >>> s = """\
... try: ... try:
... str.__nonzero__ ... str.__nonzero__
... except AttributeError: ... except AttributeError:
... pass ... pass
... """ ... """
>>> t = timeit.Timer(stmt=s) >>> timeit.timeit(stmt=s, number=100000)
>>> print "%.2f usec/pass" % (1000000 * t.timeit(number=100000)/100000) 0.9138244460009446
17.09 usec/pass >>> s = "if hasattr(str, '__bool__'): pass"
>>> s = """\ >>> timeit.timeit(stmt=s, number=100000)
... if hasattr(str, '__nonzero__'): pass 0.5829014980008651
... """ >>>
>>> t = timeit.Timer(stmt=s) >>> # attribute is present
>>> print "%.2f usec/pass" % (1000000 * t.timeit(number=100000)/100000)
4.85 usec/pass
>>> s = """\ >>> s = """\
... try: ... try:
... int.__nonzero__ ... int.__nonzero__
... except AttributeError: ... except AttributeError:
... pass ... pass
... """ ... """
>>> t = timeit.Timer(stmt=s) >>> timeit.timeit(stmt=s, number=100000)
>>> print "%.2f usec/pass" % (1000000 * t.timeit(number=100000)/100000) 0.04215312199994514
1.97 usec/pass >>> s = "if hasattr(int, '__bool__'): pass"
>>> s = """\ >>> timeit.timeit(stmt=s, number=100000)
... if hasattr(int, '__nonzero__'): pass 0.08588060699912603
... """
>>> t = timeit.Timer(stmt=s)
>>> print "%.2f usec/pass" % (1000000 * t.timeit(number=100000)/100000)
3.15 usec/pass
To give the :mod:`timeit` module access to functions you define, you can pass a To give the :mod:`timeit` module access to functions you define, you can pass a
*setup* parameter which contains an import statement:: *setup* parameter which contains an import statement::
@ -257,7 +323,5 @@ To give the :mod:`timeit` module access to functions you define, you can pass a
L.append(i) L.append(i)
if __name__ == '__main__': if __name__ == '__main__':
from timeit import Timer import timeit
t = Timer("test()", "from __main__ import test") print(timeit.timeit("test()", setup="from __main__ import test"))
print t.timeit()

2
Misc/NEWS
View file

@ -459,6 +459,8 @@ Build
Documentation Documentation
------------- -------------
- Issue #15979: Improve timeit documentation.
- Issue #16036: Improve documentation of built-in int()'s signature and - Issue #16036: Improve documentation of built-in int()'s signature and
arguments. arguments.