�
��c@s}dZd
dk
lZ
d
dklZlZ
d
dk	l
ZlZ
lZlZlZ
d
dk	lZlZlZlZ
d
dklZ
d
dkl Z!
dd	d
ddd
ddddddddddddddddddgZ"d e
d!�ed"�Z#d"eZ$ed#�Z%d$ed%�Z&d&Z'd'e'Z(d
d(k)Z)de)i*f
d)��YZ*de*f
d*��YZ+de*f
d+��YZ,d,�Z-d-d.�Z.e*�Z/e/i0Z0e/i1Z1e/i2Z2e/i3Z3e/i4Z4e/i5Z5e/i6Z6e/i7Z7e/i8Z8e/i9Z9e/i:Z:e/i;Z;e/i<Z<e/i=Z=e/i>Z>e/i?Z?e/i@Z@e/iAZAe/iBZBe/iCZCe/iDZDeEd/jo
e.�
n

d(S(0s:Random variable generators.

    integers
    --------
           uniform within range

    sequences
    ---------
           pick random element
           pick random sample
           generate random permutation

    distributions on the real line:
    ------------------------------
           uniform
           normal (Gaussian)
           lognormal
           negative exponential
           gamma
           beta
           pareto
           Weibull

    distributions on the circle (angles 0 to 2pi)
    ---------------------------------------------
           circular uniform
           von Mises

General notes on the underlying Mersenne Twister core generator:

* The period is 2**19937-1.
* It is one of the most extensively tested generators in existence.
* Without a direct way to compute N steps forward, the semantics of
  jumpahead(n) are weakened to simply jump to another distant state and rely
  on the large period to avoid overlapping sequences.
* The random() method is implemented in C, executes in a single Python step,
  and is, therefore, threadsafe.

i�(
twarn(t
MethodTypetBuiltinMethodType(tlogtexptpit
etceil(tsqrttacostcostsin(
turandom(
thexlifytRandomtseedtrandomtuniformtrandinttchoicetsamplet	randrangetshufflet
normalvariatetlognormvariatetexpovariatetvonmisesvariatetgammavariatetgausstbetavariatet
paretovariatetweibullvariatetgetstatetsetstatet	jumpaheadtWichmannHilltgetrandbitstSystemRandomig�@g@g�@i5iNcBs
eZ
dZd
Zdd�Zdd�Zd�Zd�Zd�Z	d�Z
d�Zdd	edd
e
>d�Zd�Zeed
e
>eed
�Zd�Zded�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Zd�Z RS(s�Random number generator base class used by bound module functions.

    Used to instantiate instances of Random to get generators that don't
    share state.  Especially useful for multi-threaded programs, creating
    a different instance of Random for each thread, and using the jumpahead()
    method to ensure that the generated sequences seen by each thread don't
    overlap.

    Class Random can also be subclassed if you want to use a different basic
    generator of your own devising: in that case, override the following
    methods:  random(), seed(), getstate(), setstate() and jumpahead().
    Optionally, implement a getrandombits() method so that randrange()
    can cover arbitrarily large ranges.

    icCs|i|
�
d
|_d
S(seInitialize an instance.

        Optional argument x controls seeding, as for Random.seed().
        N(RtNonet
gauss_next(tselft
x((s/sys/lib/python/random.pyt__init__Xs

cCs�|
djo[
yt
ttd
��d
�}
Wqhtj
o)



ddk}t
|i�d�}
qh
Xn

tt|�i|
�
d|_	dS(s
Initialize internal state from hashable object.

        None or no argument seeds from current time or from an operating
        system specific randomness source if available.

        If a is not None or an int or long, hash(a) is used instead.
        ii�Ni
(
R&tlongt_hexlifyt_urandomtNotImplementedErrorttimetsuperRRR'(R(t
aR/((s/sys/lib/python/random.pyRas	






c

Cs"|it
t|�i�|ifS(
s9Return internal state; can be passed to setstate() later.(tVERSIONR0RR R'(
R(((s/sys/lib/python/random.pyR tscCs`|
d
}|djo,
|
\}}|_t
t|�i|�
n
td||if��dS(s:Restore internal state from object returned by getstate().iis?state with version %s passed to Random.setstate() of version %sN(R'R0RR!t
ValueErrorR2(R(tstatetversiont
internalstate((s/sys/lib/python/random.pyR!xs




c


Cs
|i�S(
N(
R (
R(((s/sys/lib/python/random.pyt__getstate__�s
cCs|i|
�
dS(
N(
R!(R(R4((s/sys/lib/python/random.pyt__setstate__�s
c

Cs|id
|i
�fS(N((t	__class__R (
R(((s/sys/lib/python/random.pyt
__reduce__�s
i
l

cCs||
�}||
jo

td
�n

||joP
|djo6
||jo
|i
|�Sn

||i�|�Sn

td�n

||�}||jo

td�n

||}	|djoW
|	djoJ
|	|jo
|||i
|	��Sn

||||i�|	��Sn

|djo
td|||	f�n

||�}
|
|jo

td�n

|
djo
|	|
d|
}n-
|
djo
|	|
d|
}n

td�|dj
o

td�n

||jo
||
|i
|�Sn

||
||i�|�S(	s
Choose a random item from range(start, stop[, step]).

        This fixes the problem with randint() which includes the
        endpoint; in Python this is usually not what you want.
        Do not supply the 'int', 'default', and 'maxwidth' arguments.
        s!non-integer arg 1 for randrange()isempty range for randrange()s non-integer stop for randrange()i
s'empty range for randrange() (%d,%d, %d)s non-integer step for randrange()szero step for randrange()(R3t
_randbelowR(R(tstarttstoptsteptinttdefaulttmaxwidthtistarttistoptwidthtistept
n((s/sys/lib/python/random.pyR�s@























"












	





cCs|i|
|d
�S(sJReturn random integer in range [a, b], including both end points.
        i
(
R(R(R1t
b((s/sys/lib/python/random.pyR�sc
Cs�y
|i}Wnt
j
o



n}
Xt|i�|jp
t|�|joR
|d
||
dd��}||�}	x|	|
jo
||�}	qx
W|	Sn

|
|jo
td�
n

||i�|
�S(s�Return a random int in the range [0,n)

        Handles the case where n has more bits than returned
        by a single call to the underlying generator.
        gr�|
�g@sgUnderlying random() generator does not supply 
enough bits to choose from a population range this large(R$tAttributeErrorttypeRt_warn(
R(RFt_logR?t	_maxwidtht_Methodt_BuiltinMethodR$t
kt
r((s/sys/lib/python/random.pyR;�s



)







cCs|
t|i
�t|
��S(
s2Choose a random element from a non-empty sequence.(R?Rtlen(R(tseq((s/sys/lib/python/random.pyR�scCsx|djo

|i
}n

xWttd
t|
���D]:}||�|d
�}|
||
||
|<|
|<q6WdS(s�x, random=random.random -> shuffle list x in place; return None.

        Optional arg random is a 0-argument function returning a random
        float in [0.0, 1.0); by default, the standard random.random.
        i
N(R&RtreversedtxrangeRQ(R(R)RR?t
it
j((s/sys/lib/python/random.pyR�s




c
Cs�t|
�}d
|j
o

|j
n
p

t
d�n

|i}t}d	g
|}d}|djo%
|dtt|dd��7}n

||j
p
t|
d�ob
t|
�}x
t	|�D]A}	||�||	�}
||
||	<|||	d||
<q�Wn�
y�t
�}|i}xgt	|�D]Y}	||�|�}
x%|
|jo
||�|�}
qH

W||
�
|
|
||	<q,
WWnFtt
fj
o4



t|
t�o
�n

|it|
�|�Sn
X|S(
s8Chooses k unique random elements from a population sequence.

        Returns a new list containing elements from the population while
        leaving the original population unchanged.  The resulting list is
        in selection order so that all sub-slices will also be valid random
        samples.  This allows raffle winners (the sample) to be partitioned
        into grand prize and second place winners (the subslices).

        Members of the population need not be hashable or unique.  If the
        population contains repeats, then each occurrence is a possible
        selection in the sample.

        To choose a sample in a range of integers, use xrange as an argument.
        This is especially fast and space efficient for sampling from a
        large population:   sample(xrange(10000000), 60)
        issample larger than populationiiiitkeysi
N(RQR3RR?R&t_ceilRKthasattrtlistRTtsettaddt	TypeErrortKeyErrort
isinstanceRttuple(
R(t
populationRORFRt_inttresulttsetsizetpoolRURVtselectedtselected_add((s/sys/lib/python/random.pyR
s>%



	






%






	
	












cCs|
||
|i�S(
s(Get a random number in the range [a, b).(
R(R(R1RG((s/sys/lib/python/random.pyRR
scCsj|i}xR|�}d
|�}t
|d|}||d}|t|�j
o
Pq
q|
||S(s\Normal distribution.

        mu is the mean, and sigma is the standard deviation.

        g��@(Rt
NV_MAGICCONSTRK(R(tmutsigmaRtu1tu2t
ztzz((s/sys/lib/python/random.pyRX
s
	

	






cCst|i
|
|��S(
s�Log normal distribution.

        If you take the natural logarithm of this distribution, you'll get a
        normal distribution with mean mu and standard deviation sigma.
        mu can have any value, and sigma must be greater than zero.

        (t_expR(R(RiRj((s/sys/lib/python/random.pyRq
scCs?|i}|�}x|d
j
o

|�}q
Wt
|�|
S(s�Exponential distribution.

        lambd is 1.0 divided by the desired mean.  (The parameter would be
        called "lambda", but that is a reserved word in Python.)  Returned
        values range from 0 to positive infinity.

        gH���>(RRK(R(tlambdRt
u((s/sys/lib/python/random.pyR}
s
	
	


cCs:
|i}|d
j
o
t
|�Sn

dtdd||�}|td|�d|}d||d|}x~|�}tt|�}d||||}	|||	}
|�}||
d|
jp
||
td|
�j
o
Pqv
qv|�}|djo
|
t
t|	�}
n
|
t
t|	�}
|
S(sF
Circular data distribution.

        mu is the mean angle, expressed in radians between 0 and 2*pi, and
        kappa is the concentration parameter, which must be greater than or
        equal to zero.  If kappa is equal to zero, this distribution reduces
        to a uniform random angle over the range 0 to 2*pi.

        g���g�@g@g�RtTWOPIt_sqrtt_cost_piRot_acos(R(RitkappaRR1RGRPRkRmt
ft
cRltu3ttheta((s/sys/lib/python/random.pyR�s&	





	

	0
	


cCs%|
d
j
p

|d
j
o

td�n

|i
}|
djo�td|
d�}|
t}|
|}x�|�}d|jo

djn
p
qhn

d|�}t|d|�|}	|
t|	�}
|||}|||	|
}|td|d
jp
|t|�jo
|
|Sqh
qhn�
djo:
|�}
x|
dj
o

|�}
qC

Wt|
�|Sn�
x�|�}
t|
t}||
}|dj
o
|d|
}
n
t|||
�}
|�}|djo
||
|
dj
o
Pq
qt

|t|
�j
o
Pqt

qt
|
|SdS(	sxGamma distribution.  Not the gamma function!

        Conditions on the parameters are alpha > 0 and beta > 0.

        gs*gammavariate: alpha and beta must be > 0.0g�@gH���>g�P��@N(R3RRstLOG4RKRot
SG_MAGICCONSTt_e(R(talphatbetaRtainvtbbbtcccRkRlt
vR)RmRPRqRGt
p((s/sys/lib/python/random.pyR�sJ

	






	








,

	



	






	



	


cCs�|i}|i
}d|_
|djoQ
|�t}td
td|���}t|�|}t|�||_
n

|
||S(s�Gaussian distribution.

        mu is the mean, and sigma is the standard deviation.  This is
        slightly faster than the normalvariate() function.

        Not thread-safe without a lock around calls.

        g��RR'R&RrRsRKRtt_sin(R(RiRjRRmtx2pitg2rad((s/sys/lib/python/random.pyRs	
	
	






cCsC|i|
d
�}|djo
dSn
|||i|d
�SdS(s�Beta distribution.

        Conditions on the parameters are alpha > 0 and beta > 0.
        Returned values range between 0 and 1.

        g�gN(
R(R(RR�t
y((s/sys/lib/python/random.pyR9s



cCs%d
|i�}d
t
|d
|
�S(s3Pareto distribution.  alpha is the shape parameter.g�Rtpow(R(RRq((s/sys/lib/python/random.pyRKs
cCs,d
|i�}|
t
t|�d
|�S(sfWeibull distribution.

        alpha is the scale parameter and beta is the shape parameter.

        g�RR�RK(R(RR�Rq((s/sys/lib/python/random.pyRTs
N(!t__name__t
__module__t__doc__R2R&R*RR R!R7R8R:R?tBPFRRRKt_MethodTypet_BuiltinMethodTypeR;RRRRRRRRRRRRR(((s/sys/lib/python/random.pyREs6						
?	

		J					2	B	5			cBs\eZ
dZd	d
�Zd�Zd�Zd�Zd�Zdddd�Z	d	d�Z
RS(
i
cCs

|
djo[
yt
ttd
��d
�}
Wqhtj
o)



ddk}t
|i�d�}
qh
Xn

t|
tt
f�p
t|
�}
n

t	|
d�\}
}t	|
d�\}
}t	|
d�\}
}t|�dt|�dt|�df|_
d|_dS(	s�Initialize internal state from hashable object.

        None or no argument seeds from current time or from an operating
        system specific randomness source if available.

        If a is not None or an int or long, hash(a) is used instead.

        If a is an int or long, a is used directly.  Distinct values between
        0 and 27814431486575L inclusive are guaranteed to yield distinct
        internal states (this guarantee is specific to the default
        Wichmann-Hill generator).
        ii�Ni
i<vibvirvi
(R&R+R,R-R.R/R_R?thashtdivmodt_seedR'(R(R1R/R)R�Rm((s/sys/lib/python/random.pyRes









0c
Csj|i\}
}}d
|
d}
d|d}d|d}|
||f|_|
d|d|d	d
S(s3Get the next random number in the range [0.0, 1.0).i�i=vi�icvi�isvg@�@g��g��g�R�(R(R)R�Rm((s/sys/lib/python/random.pyR�s



c

Cs|i|i
|ifS(
s9Return internal state; can be passed to setstate() later.(R2R�R'(
R(((s/sys/lib/python/random.pyR �scCsM|
d
}|djo
|
\}|_|_
n
td||if��dS(s:Restore internal state from object returned by getstate().ii
s?state with version %s passed to Random.setstate() of version %sN(R�R'R3R2(R(R4R5((s/sys/lib/python/random.pyR!�s




cCs�|
d
jp
td��n

|i
\}}}t|td|
d��d}t|td|
d��d}t|td|
d��d}|||f|_
d	S(
s�Act as if n calls to random() were made, but quickly.

        n is an int, greater than or equal to 0.

        Example use:  If you have 2 threads and know that each will
        consume no more than a million random numbers, create two Random
        objects r1 and r2, then do
            r2.setstate(r1.getstate())
            r2.jumpahead(1000000)
        Then r1 and r2 will use guaranteed-disjoint segments of the full
        period.
        isn must be >= 0i�i=vi�icvi�isvN(R3R�R?R�(R(RFR)R�Rm((s/sys/lib/python/random.pyR"�s



 
 
 
icCs�t|
�t|�jo
t|�jo

t
jn
p
td
��n

d|
j
o

djn
o9
d|j
o

djn
o
d|j
o

djn
p
td��n

d|
jo
|jo

|jn
o}
ddk}t|i�d�}t
|d@|d?A�}t|d�\}}
t|d�\}}t|d�\}}n

|
p
d	|p
d	|p
d	f|_d|_	dS(
sjSet the Wichmann-Hill seed from (x, y, z).

        These must be integers in the range [0, 256).
        sseeds must be integersii
sseeds must be in range(0, 256)i�Ni�ii
(
RIR?R]R3R/R+R�R�R&R'(R(R)R�RmR/t
t((s/sys/lib/python/random.pyt__whseed�s<

Z

*




'cCs�|
d
jo
|i
�
d
Sn

t|
�}
t|
d�\}
}t|
d�\}
}t|
d�\}
}||
dp
d}||
dp
d}||
dp
d}|i
|||�
d
S(sb
Seed from hashable object's hash code.

        None or no argument seeds from current time.  It is not guaranteed
        that objects with distinct hash codes lead to distinct internal
        states.

        This is obsolete, provided for compatibility with the seed routine
        used prior to Python 2.1.  Use the .seed() method instead.
        Ni
i
(R&t_WichmannHill__whseedR�R�(R(R1R)R�Rm((s/sys/lib/python/random.pytwhseed�s











N(R�R�R2R&RRR R!R"R�R�(((s/sys/lib/python/random.pyR#as			
	cBsFeZ
dZd
�Zd�Zd�ZeZZd�ZeZ	Z
RS(s�Alternate random number generator using sources provided
    by the operating system (such as /dev/urandom on Unix or
    CryptGenRandom on Windows).

     Not available on all systems (see os.urandom() for details).
    c

Cs!tt
td
��d�d?tS(s3Get the next random number in the range [0.0, 1.0).iii(R+R,R-t	RECIP_BPF(
R(((s/sys/lib/python/random.pyR�scCsy|
d
j
o
td��n

|
t
|
�jo
td��n

|
dd}ttt|��d�}||d|
?S(s>getrandbits(k) -> x.  Generates a long int with k random bits.is(number of bits must be greater than zeros#number of bits should be an integeriii(R3R?R]R+R,R-(R(ROtbytesR)((s/sys/lib/python/random.pyR$s






c

Osd
S(s<Stub method.  Not used for a system random number generator.N(
R&(R(targstkwds((s/sys/lib/python/random.pyt_stub
sc
Ostd
��dS(sAMethod should not be called for a system random number generator.s*System entropy source does not have state.N(
R.(R(R�R�((s/sys/lib/python/random.pyt_notimplementeds(R�R�R�RR$R�RR"R�R R!(((s/sys/lib/python/random.pyR%�s		
	
	cCs�d
dk}|GdG|
i
GHd}d}d}d}|i�}xVt|�D]H}	|
|�}
||
7}||
|
}t|
|�}t|
|�}qMW|i�}t||d�GdG||}t||||�}
d||
||fGHdS(	Ni�ttimesgg _�Bg _��ssec,s!avg %g, stddev %g, min %g, max %g(R/R�trangetmintmaxtroundRs(RFtfuncR�R/ttotaltsqsumtsmallesttlargesttt0RUR)tt1tavgtstddev((s/sys/lib/python/random.pyt_test_generators(





















i�c

Cs�t|t
d�
t|td�
t|td
�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
t|td�
dS(Ngg���?g����g@g�����4@gi@g@((gg�gg�gg�g{��?g�g����g�g����g@(g��g�����g��g@g�g4@g�gi@g�gg�g@g@(R�RRRRRRR(
t
N((s/sys/lib/python/random.pyt_test/s














t__main__(FR�twarningsRRJttypesR
R�RR�tmathRRKRRoRRuRR~RRXRRsR	RvR
RtRR�tosRR-tbinasciiR
R,t__all__RhRrR|R}R�R�t_randomRR#R%R�R�t_instRRRRRRRRRRRRRRRRRR R!R"R$R�(((s/sys/lib/python/random.pys<module>'sd

(
"



	


	





��"