Commit Graph

76 Commits

Author SHA1 Message Date
Tim Peters 0e1159583c shuffle() doscstring: Removed warning about sequence length
versus generator period.  While this was a real weakness of the
older WH generator for lists with just a few dozen elements,
and so could potentially bite the naive ;-), the Twister should
show excellent behavior up to at least 600 elements.

Module docstring:  reflowed some jarringly short lines.
2006-06-10 22:51:45 +00:00
Tim Peters c17976e983 Another crack at bug #1460340: make random.sample(dict)
work, this time by ugly brute force.
2006-04-01 00:26:53 +00:00
Raymond Hettinger 3c3346daa9 SF bug #1460340: random.sample can raise KeyError
Fix the hit and miss style of testing for sets and dicts.
2006-03-29 09:13:13 +00:00
Tim Peters 9e34c04732 Whitespace normalization (via reindent.py). 2005-08-26 15:20:46 +00:00
Raymond Hettinger 91e27c253c Implement random.sample() using sets instead of dicts. 2005-08-19 01:36:35 +00:00
Raymond Hettinger 42406e6f27 SF patch #1191489: Simplify logic in random.py 2005-04-30 09:02:51 +00:00
Raymond Hettinger 362b929222 Removed unused line. 2005-01-03 07:33:16 +00:00
Raymond Hettinger ffdb8bb99c Use floor division operator. 2004-09-27 15:29:05 +00:00
Raymond Hettinger 23f1241dc6 SF #1027105: HardwareRandom should be renamed OSRandom
Renamed the new generator at Trevor's recommendation.
The name HardwareRandom suggested a bit more than it
delivered (no radioactive decay detectors or such).
2004-09-13 22:23:21 +00:00
Raymond Hettinger c1c43cad63 Fulfill Martin's request to use try/except rather than a "look before
you leap" approach.  Makes the early call to os.urandom() unnecessary.
2004-09-05 00:00:42 +00:00
Raymond Hettinger bdbdfd731d SF bug #1022010: Import random fails
* trap NotImplementedError raised by os.urandom calls when not available
  on a particular system.
2004-09-03 19:09:22 +00:00
Tim Peters 7c2a85b2d4 HardwareRandom: Go back to multiplying by 2**-BPF instead of using
ldexp.  Both methods are exact, and return the same results.  Turns out
multiplication is a few (but just a few) percent faster on my box.
They're both significantly faster than using struct with a Q format
to convert bytes to a 64-bit long (struct.unpack() appears to lose due
to the tuple creation/teardown overhead), and calling _hexlify is
significantly faster than doing bytes.encode('hex').  So we appear to
have hit a local minimum (wrt speed) here.
2004-08-31 02:19:55 +00:00
Raymond Hettinger 3fa19d7ff8 Take advantage of the math library's ldexp for assembling a float by
components without division and without roundoff error for properly
sized mantissas (i.e. on systems with 53 or more mantissa bits per
float).  Eliminates the previous implementation's rounding bias as
aptly demonstrated by Tim Peters.
2004-08-31 01:05:15 +00:00
Raymond Hettinger 356a4599ac Teach the random module about os.urandom().
* Use it for seeding when it is available.
* Provide an alternate generator based on it.
2004-08-30 06:14:31 +00:00
Raymond Hettinger 5dae505bbd Fix typo in comment. 2004-06-07 02:07:15 +00:00
Raymond Hettinger 9021c95595 SF bug #964230: random.choice([]) should return more intelligible exception
Add a comment to make the traceback less mysterious.
2004-06-05 14:53:22 +00:00
Jeremy Hylton 2b55d35850 Reflow long line. 2004-02-23 17:27:57 +00:00
Tim Peters 58eb11cf62 Whitespace normalization. 2004-01-18 20:29:55 +00:00
Raymond Hettinger 85c20a41df Implement and apply PEP 322, reverse iteration 2003-11-06 14:06:48 +00:00
Raymond Hettinger 411c602349 Minor fixup. "Random" was listed twice in __all__. 2003-10-12 17:14:11 +00:00
Raymond Hettinger 2f726e9093 SF bug #812202: randint is always even
* Added C coded getrandbits(k) method that runs in linear time.
* Call the new method from randrange() for ranges >= 2**53.
* Adds a warning for generators not defining getrandbits() whenever they
  have a call to randrange() with too large of a population.
2003-10-05 09:09:15 +00:00
Raymond Hettinger 66d09f1b30 SF bug #801342: Bug (documentation or real, your choice) in random.sample.
random.sample() uses one of two algorithms depending on the ratio of the
sample size to the population size.  One of the algorithms accepted any
iterable population argument so long as it defined __len__().  The other
had a stronger requirement that the population argument be indexable.

While it met the documentation specifications which insisted that the
population argument be a sequence, it made random.sample() less usable
with sets.  So, the second algorithm was modified to coerce non-indexable
iterables and dictionaries into a tuple before proceeding.
2003-09-06 04:25:54 +00:00
Raymond Hettinger 6229713221 Eliminate unnecessary eval() in tests 2003-08-30 01:24:19 +00:00
Raymond Hettinger 3081d59f92 SF bug #778964: bad seed in python 2.3 random
The default seed is time.time().
Multiplied by 256 before truncating so that fractional seconds are used.
This way, two successive calls to random.seed() are much more likely
to produce different sequences.
2003-08-09 18:30:57 +00:00
Raymond Hettinger f8a52d38ad Removed deprecated functions 2003-08-05 12:23:19 +00:00
Raymond Hettinger 5f078ff7f0 SF bug #759889: Pickling of Random is broken
* Implement __reduce__() to support pickling.
* Add a test case to prove a successful roundtrip through pickle.
2003-06-24 20:29:04 +00:00
Tim Peters 76ca1d428f randrange(): Repaired my overly optimistic rewrite, and added comments
explaining what's wrong with the two simpler variants.
2003-06-19 03:46:46 +00:00
Tim Peters afb8979771 randrange(): 2.3 can no longer raises OverflowError on an int() call, so
some of this code because useless, and (worse) could return a long
instead of int (in Zope that's important, because a long can't be used
as a key in an IOBTree or IIBTree).
2003-06-19 03:23:06 +00:00
Raymond Hettinger fdbe5223b7 SF bug #753602: random.sample not properly documented
The docs were fine but the "int=int" in the function call was both
ugly and confusing.  Moved it inside the body of the function definition.
2003-06-13 07:01:51 +00:00
Raymond Hettinger b98154e424 'sum' is no longer a good variable name. Use 'total' instead. 2003-05-24 17:26:02 +00:00
Raymond Hettinger 7b0cf76b72 * Migrate sample distribution test from random.py to test_random.py.
* Use Sets module to more clearly articulate a couple of tests.
2003-01-17 17:23:23 +00:00
Raymond Hettinger 145a4a0f10 Much clearer when super() is used. 2003-01-07 10:25:55 +00:00
Raymond Hettinger 3dd990c53a Move the statistical tests for four distributions into the unittest suite. 2003-01-05 09:20:06 +00:00
Raymond Hettinger 73ced7ee99 Correct long standing bugs in the methods for random distributions.
The range of u=random() is [0,1), so log(u) and 1/x can fail.
Fix by setting u=1-random() or by reselecting for a usable value.

Will backport.
2003-01-04 09:26:32 +00:00
Raymond Hettinger 8b9aa8dbba Remove the random=None nonsense from sample() before it gets set in stone.
It was once available so that faster generators could be substituted.  Now,
that is less necessary and preferrably done via subclassing.

Also, clarified and shortened the comments for sample().
2003-01-04 05:20:33 +00:00
Raymond Hettinger 40f6217092 SF patch 658251: Install a C implementation of the Mersenne Twister as the
core generator for random.py.
2002-12-29 23:03:38 +00:00
Raymond Hettinger c1ac194e94 Clarify and speedup test. 2002-12-07 09:25:05 +00:00
Raymond Hettinger 311f419628 Improve comments. Clarify docs.
Replace "type(0)" with "int".
Replace "while 1" with "while True"
2002-11-18 09:01:24 +00:00
Raymond Hettinger c0b4034b81 Improved clarity and thoroughness of docstring.
Added design notes in comments.
Used better variable names.
Eliminated the unsavory "pool[-k:]" which was an aspiring bug (for k==0).
Used if/else to show the two algorithms in parallel style.
Added one more test assertion.
2002-11-13 15:26:37 +00:00
Andrew M. Kuchling ea002a1aff Docstring typo fix 2002-11-13 13:25:46 +00:00
Raymond Hettinger f24eb35d18 SF patch 629637: Add sample(population, k) method to the random module.
Used for random sampling without replacement.
2002-11-12 17:41:57 +00:00
Tim Peters 9146f27b77 SF bug 594996: OverflowError in random.randrange
Loosened the acceptable 'start' and 'stop' arguments so that any
Python (bounded) ints can be used.  So, e.g., randrange(-sys.maxint-1,
sys.maxint) no longer blows up.
2002-08-16 03:41:39 +00:00
Raymond Hettinger ef4d4bdc3c Whitespace normalized. 2002-05-23 23:58:17 +00:00
Raymond Hettinger c32f0336e0 Deprecated Random.cunifvariate clearing bug 506647. Also, added docstrings. 2002-05-23 19:44:49 +00:00
Tim Peters 8ac1495a6a Whitespace normalization. 2002-05-23 15:15:30 +00:00
Guido van Rossum 570764ddce Use spaces, not tabs for indents.
Raymond Hettinger, beware!  This caused the Unix install to fail!
2002-05-14 14:08:12 +00:00
Raymond Hettinger b760efb08d Closes patch 529408 deprecating random.stdgamma(). 2002-05-14 06:40:34 +00:00
Raymond Hettinger ca6cdc2c02 Closes SF bug 527139. Moved temp variables inside guard. 2002-05-13 23:40:14 +00:00
Tim Peters 46c04e140c random.gauss() uses a piece of hidden state used by nothing else,
and the .seed() and .whseed() methods failed to reset it.  In other
words, setting the seed didn't completely determine the sequence of
results produced by random.gauss().  It does now.  Programs repeatedly
mixing calls to a seed method with calls to gauss() may see different
results now.

Bugfix candidate (random.gauss() has always been broken in this way),
despite that it may change results.
2002-05-05 20:40:00 +00:00
Neal Norwitz f42d32cb27 BDFL agreed with Tim: rehabilitate randint(). 2002-04-13 14:41:19 +00:00