1 \" Copyright (c) 1993 Martin Birgmeier
2 .\" All rights reserved.
4 .\" You may redistribute unmodified or modified versions of this source
5 .\" code provided that the above copyright notice and this and the
6 .\" following conditions are retained.
8 .\" This software is provided ``as is'', and comes with no warranties
9 .\" of any kind. I shall in no event be liable for anything that happens
10 .\" to anyone/anything when using this software.
12 .\" @(#)rand48.3 V1.0 MB 8 Oct 1993
28 .Nd pseudo random number generators and initialization routines
30 .Fd #include <stdlib.h>
34 .Fn erand48 "unsigned short xseed[3]"
38 .Fn nrand48 "unsigned short xseed[3]"
42 .Fn jrand48 "unsigned short xseed[3]"
44 .Fn srand48 "long seed"
45 .Ft "unsigned short *"
46 .Fn seed48 "unsigned short xseed[3]"
48 .Fn lcong48 "unsigned short p[7]"
52 family of functions generates pseudo-random numbers using a linear
53 congruential algorithm working on integers 48 bits in size. The
54 particular formula employed is
55 r(n+1) = (a * r(n) + c) mod m
56 where the default values are
57 for the multiplicand a = 0xfdeece66d = 25214903917 and
58 the addend c = 0xb = 11. The modulo is always fixed at m = 2 ** 48.
59 r(n) is called the seed of the random number generator.
61 For all the six generator routines described next, the first
62 computational step is to perform a single iteration of the algorithm.
67 return values of type double. The full 48 bits of r(n+1) are
68 loaded into the mantissa of the returned value, with the exponent set
69 such that the values produced lie in the interval [0.0, 1.0).
74 return values of type long in the range
75 [0, 2**31-1]. The high-order (31) bits of
76 r(n+1) are loaded into the lower bits of the returned value, with
77 the topmost (sign) bit set to zero.
82 return values of type long in the range
83 [-2**31, 2**31-1]. The high-order (32) bits of
84 r(n+1) are loaded into the returned value.
90 use an internal buffer to store r(n). For these functions
91 the initial value of r(0) = 0x1234abcd330e = 20017429951246.
98 use a user-supplied buffer to store the seed r(n),
99 which consists of an array of 3 shorts, where the zeroth member
100 holds the least significant bits.
102 All functions share the same multiplicand and addend.
105 is used to initialize the internal buffer r(n) of
110 such that the 32 bits of the seed value are copied into the upper 32 bits
111 of r(n), with the lower 16 bits of r(n) arbitrarily being set to 0x330e.
112 Additionally, the constant multiplicand and addend of the algorithm are
113 reset to the default values given above.
116 also initializes the internal buffer r(n) of
121 but here all 48 bits of the seed can be specified in an array of 3 shorts,
122 where the zeroth member specifies the lowest bits. Again,
123 the constant multiplicand and addend of the algorithm are
124 reset to the default values given above.
126 returns a pointer to an array of 3 shorts which contains the old seed.
127 This array is statically allocated, thus its contents are lost after
133 allows full control over the multiplicand and addend used in
146 An array of 7 shorts is passed as parameter; the first three shorts are
147 used to initialize the seed; the second three are used to initialize the
148 multiplicand; and the last short is used to initialize the addend.
149 It is thus not possible to use values greater than 0xffff as the addend.
151 Note that all three methods of seeding the random number generator
152 always also set the multiplicand and addend for any of the six
155 For a more powerful random number generator, see