2
:mod:`random` --- Generate pseudo-random numbers
3
================================================
6
:synopsis: Generate pseudo-random numbers with various common distributions.
9
This module implements pseudo-random number generators for various
12
For integers, uniform selection from a range. For sequences, uniform selection
13
of a random element, a function to generate a random permutation of a list
14
in-place, and a function for random sampling without replacement.
16
On the real line, there are functions to compute uniform, normal (Gaussian),
17
lognormal, negative exponential, gamma, and beta distributions. For generating
18
distributions of angles, the von Mises distribution is available.
20
Almost all module functions depend on the basic function :func:`random`, which
21
generates a random float uniformly in the semi-open range [0.0, 1.0). Python
22
uses the Mersenne Twister as the core generator. It produces 53-bit precision
23
floats and has a period of 2\*\*19937-1. The underlying implementation in C is
24
both fast and threadsafe. The Mersenne Twister is one of the most extensively
25
tested random number generators in existence. However, being completely
26
deterministic, it is not suitable for all purposes, and is completely unsuitable
27
for cryptographic purposes.
29
The functions supplied by this module are actually bound methods of a hidden
30
instance of the :class:`random.Random` class. You can instantiate your own
31
instances of :class:`Random` to get generators that don't share state. This is
32
especially useful for multi-threaded programs, creating a different instance of
33
:class:`Random` for each thread, and using the :meth:`jumpahead` method to make
34
it likely that the generated sequences seen by each thread don't overlap.
36
Class :class:`Random` can also be subclassed if you want to use a different
37
basic generator of your own devising: in that case, override the :meth:`random`,
38
:meth:`seed`, :meth:`getstate`, :meth:`setstate` and :meth:`jumpahead` methods.
39
Optionally, a new generator can supply a :meth:`getrandbits` method --- this
40
allows :meth:`randrange` to produce selections over an arbitrarily large range.
43
the :meth:`getrandbits` method.
45
As an example of subclassing, the :mod:`random` module provides the
46
:class:`WichmannHill` class that implements an alternative generator in pure
47
Python. The class provides a backward compatible way to reproduce results from
48
earlier versions of Python, which used the Wichmann-Hill algorithm as the core
49
generator. Note that this Wichmann-Hill generator can no longer be recommended:
50
its period is too short by contemporary standards, and the sequence generated is
51
known to fail some stringent randomness tests. See the references below for a
52
recent variant that repairs these flaws.
54
.. versionchanged:: 2.3
55
Substituted MersenneTwister for Wichmann-Hill.
57
Bookkeeping functions:
60
.. function:: seed([x])
62
Initialize the basic random number generator. Optional argument *x* can be any
63
:term:`hashable` object. If *x* is omitted or ``None``, current system time is used;
64
current system time is also used to initialize the generator when the module is
65
first imported. If randomness sources are provided by the operating system,
66
they are used instead of the system time (see the :func:`os.urandom` function
67
for details on availability).
69
.. versionchanged:: 2.4
70
formerly, operating system resources were not used.
72
If *x* is not ``None`` or an int or long, ``hash(x)`` is used instead. If *x* is
73
an int or long, *x* is used directly.
76
.. function:: getstate()
78
Return an object capturing the current internal state of the generator. This
79
object can be passed to :func:`setstate` to restore the state.
83
.. versionchanged:: 2.6
84
State values produced in Python 2.6 cannot be loaded into earlier versions.
87
.. function:: setstate(state)
89
*state* should have been obtained from a previous call to :func:`getstate`, and
90
:func:`setstate` restores the internal state of the generator to what it was at
91
the time :func:`setstate` was called.
96
.. function:: jumpahead(n)
98
Change the internal state to one different from and likely far away from the
99
current state. *n* is a non-negative integer which is used to scramble the
100
current state vector. This is most useful in multi-threaded programs, in
101
conjunction with multiple instances of the :class:`Random` class:
102
:meth:`setstate` or :meth:`seed` can be used to force all instances into the
103
same internal state, and then :meth:`jumpahead` can be used to force the
104
instances' states far apart.
106
.. versionadded:: 2.1
108
.. versionchanged:: 2.3
109
Instead of jumping to a specific state, *n* steps ahead, ``jumpahead(n)``
110
jumps to another state likely to be separated by many steps.
113
.. function:: getrandbits(k)
115
Returns a python :class:`long` int with *k* random bits. This method is supplied
116
with the MersenneTwister generator and some other generators may also provide it
117
as an optional part of the API. When available, :meth:`getrandbits` enables
118
:meth:`randrange` to handle arbitrarily large ranges.
120
.. versionadded:: 2.4
122
Functions for integers:
125
.. function:: randrange([start,] stop[, step])
127
Return a randomly selected element from ``range(start, stop, step)``. This is
128
equivalent to ``choice(range(start, stop, step))``, but doesn't actually build a
131
.. versionadded:: 1.5.2
134
.. function:: randint(a, b)
136
Return a random integer *N* such that ``a <= N <= b``.
138
Functions for sequences:
141
.. function:: choice(seq)
143
Return a random element from the non-empty sequence *seq*. If *seq* is empty,
144
raises :exc:`IndexError`.
147
.. function:: shuffle(x[, random])
149
Shuffle the sequence *x* in place. The optional argument *random* is a
150
0-argument function returning a random float in [0.0, 1.0); by default, this is
151
the function :func:`random`.
153
Note that for even rather small ``len(x)``, the total number of permutations of
154
*x* is larger than the period of most random number generators; this implies
155
that most permutations of a long sequence can never be generated.
158
.. function:: sample(population, k)
160
Return a *k* length list of unique elements chosen from the population sequence.
161
Used for random sampling without replacement.
163
.. versionadded:: 2.3
165
Returns a new list containing elements from the population while leaving the
166
original population unchanged. The resulting list is in selection order so that
167
all sub-slices will also be valid random samples. This allows raffle winners
168
(the sample) to be partitioned into grand prize and second place winners (the
171
Members of the population need not be :term:`hashable` or unique. If the population
172
contains repeats, then each occurrence is a possible selection in the sample.
174
To choose a sample from a range of integers, use an :func:`xrange` object as an
175
argument. This is especially fast and space efficient for sampling from a large
176
population: ``sample(xrange(10000000), 60)``.
178
The following functions generate specific real-valued distributions. Function
179
parameters are named after the corresponding variables in the distribution's
180
equation, as used in common mathematical practice; most of these equations can
181
be found in any statistics text.
184
.. function:: random()
186
Return the next random floating point number in the range [0.0, 1.0).
189
.. function:: uniform(a, b)
191
Return a random floating point number *N* such that ``a <= N < b`` for
192
``a <= b`` and ``b <= N < a`` for ``b < a``.
195
.. function:: triangular(low, high, mode)
197
Return a random floating point number *N* such that ``low <= N < high`` and
198
with the specified *mode* between those bounds. The *low* and *high* bounds
199
default to zero and one. The *mode* argument defaults to the midpoint
200
between the bounds, giving a symmetric distribution.
202
.. versionadded:: 2.6
205
.. function:: betavariate(alpha, beta)
207
Beta distribution. Conditions on the parameters are ``alpha > 0`` and ``beta >
208
0``. Returned values range between 0 and 1.
211
.. function:: expovariate(lambd)
213
Exponential distribution. *lambd* is 1.0 divided by the desired mean. (The
214
parameter would be called "lambda", but that is a reserved word in Python.)
215
Returned values range from 0 to positive infinity.
218
.. function:: gammavariate(alpha, beta)
220
Gamma distribution. (*Not* the gamma function!) Conditions on the parameters
221
are ``alpha > 0`` and ``beta > 0``.
224
.. function:: gauss(mu, sigma)
226
Gaussian distribution. *mu* is the mean, and *sigma* is the standard deviation.
227
This is slightly faster than the :func:`normalvariate` function defined below.
230
.. function:: lognormvariate(mu, sigma)
232
Log normal distribution. If you take the natural logarithm of this
233
distribution, you'll get a normal distribution with mean *mu* and standard
234
deviation *sigma*. *mu* can have any value, and *sigma* must be greater than
238
.. function:: normalvariate(mu, sigma)
240
Normal distribution. *mu* is the mean, and *sigma* is the standard deviation.
243
.. function:: vonmisesvariate(mu, kappa)
245
*mu* is the mean angle, expressed in radians between 0 and 2\*\ *pi*, and *kappa*
246
is the concentration parameter, which must be greater than or equal to zero. If
247
*kappa* is equal to zero, this distribution reduces to a uniform random angle
248
over the range 0 to 2\*\ *pi*.
251
.. function:: paretovariate(alpha)
253
Pareto distribution. *alpha* is the shape parameter.
256
.. function:: weibullvariate(alpha, beta)
258
Weibull distribution. *alpha* is the scale parameter and *beta* is the shape
262
Alternative Generators:
264
.. class:: WichmannHill([seed])
266
Class that implements the Wichmann-Hill algorithm as the core generator. Has all
267
of the same methods as :class:`Random` plus the :meth:`whseed` method described
268
below. Because this class is implemented in pure Python, it is not threadsafe
269
and may require locks between calls. The period of the generator is
270
6,953,607,871,644 which is small enough to require care that two independent
271
random sequences do not overlap.
274
.. function:: whseed([x])
276
This is obsolete, supplied for bit-level compatibility with versions of Python
277
prior to 2.1. See :func:`seed` for details. :func:`whseed` does not guarantee
278
that distinct integer arguments yield distinct internal states, and can yield no
279
more than about 2\*\*24 distinct internal states in all.
282
.. class:: SystemRandom([seed])
284
Class that uses the :func:`os.urandom` function for generating random numbers
285
from sources provided by the operating system. Not available on all systems.
286
Does not rely on software state and sequences are not reproducible. Accordingly,
287
the :meth:`seed` and :meth:`jumpahead` methods have no effect and are ignored.
288
The :meth:`getstate` and :meth:`setstate` methods raise
289
:exc:`NotImplementedError` if called.
291
.. versionadded:: 2.4
293
Examples of basic usage::
295
>>> random.random() # Random float x, 0.0 <= x < 1.0
297
>>> random.uniform(1, 10) # Random float x, 1.0 <= x < 10.0
299
>>> random.randint(1, 10) # Integer from 1 to 10, endpoints included
301
>>> random.randrange(0, 101, 2) # Even integer from 0 to 100
303
>>> random.choice('abcdefghij') # Choose a random element
306
>>> items = [1, 2, 3, 4, 5, 6, 7]
307
>>> random.shuffle(items)
309
[7, 3, 2, 5, 6, 4, 1]
311
>>> random.sample([1, 2, 3, 4, 5], 3) # Choose 3 elements
318
M. Matsumoto and T. Nishimura, "Mersenne Twister: A 623-dimensionally
319
equidistributed uniform pseudorandom number generator", ACM Transactions on
320
Modeling and Computer Simulation Vol. 8, No. 1, January pp.3-30 1998.
322
Wichmann, B. A. & Hill, I. D., "Algorithm AS 183: An efficient and portable
323
pseudo-random number generator", Applied Statistics 31 (1982) 188-190.
added
removed
Doc/library/random.rst
