/* $Id$ Part of SWI-Prolog Author: R.A. O'Keefe, V.S. Costa, L. Damas, Jan Wielemaker E-mail: wielemak@science.uva.nl WWW: http://www.swi-prolog.org Copyright (C): Universidade do Porto, University of Amsterdam */ /************************************************************************* * * * YAP Prolog * * * * Yap Prolog was developed at NCCUP - Universidade do Porto * * * * Copyright L.Damas, V.S.Costa and Universidade do Porto 1985-1997 * * * ************************************************************************** * * * File: random.yap * * Last rev: 5/12/99 * * mods: * * comments: Random operations * * * *************************************************************************/ % original code from RA O'Keefe. % This is algorithm AS 183 from Applied Statistics. I also have a C % version. It is really very good. It is straightforward to make a % version which yields 15-bit random integers using only integer % arithmetic. :- module(random, [ random/1, % -float random/3, % +Low, +High, -Random randseq/3, % +Size, +Max, -Set randset/3, % +Size, +Max, -List getrand/1, % -State setrand/1 % +State ]). :- use_module(library(pairs)). :- use_foreign_library(foreign(random)). /** Random numbers Random number generator developed as part of the DEC10 library. The algorithm is based on AS 183 from Applied Statistics. Originally implemented ib Prolog by Richard O'Keeke. The SWI-Prolog versions is a translation of a C-version for YAP based on the orginal source. @copyright DEC10 version: Public domain, YAP: Artistic @author R.A. O'Keefe, V.S. Costa, L. Damas, Jan Wielemaker @see Built-in function random/1: A is random(10) */ /******************************* * C PRIMITIVES * *******************************/ %% random(-R:float) is det. % % Binds R to a new random number in [0.0,1.0). % % @see setrand/1, getrand/1. %% setrand(+State:rand(A,B,C)) is det. %% getrand(+State:rand(A,B,C)) is det. % % Query/set the state of the random library. A, B and C are % integers in the range 1..30,000. The initial state is predefined % and can be extracted using getrand/1: % % == % ?- getrand(X). % X = rand(27314, 9213, 17773). % == % % @see random/1. /******************************* * PROLOG * *******************************/ %% random(+L:int, +U:int, -R:int) is det. %% random(+L:float, +U:float, -R:float) is det. % % Binds R to a random number in [L,U). If L and U are both % integers, R is an integer, Otherwise, R is a float. Note that U % will *never* be generated. % % @bug The state is only 48-bits. This is insufficient for % generating uniformely distributed integers in a very % large domain. random(L, U, R) :- integer(L), integer(U), !, random(X), R is L+floor((U-L)*X). random(L, U, R) :- number(L), number(U), !, random(X), R is L+((U-L)*X). %% randset(+K:int, +N:int, -S:list(int)) is det. % % S is a sorted list of K integers in the range 1..N. % % @see randseq/3. randset(K, N, S) :- K >= 0, K =< N, randset(K, N, [], S). randset(0, _, S, S) :- !. randset(K, N, Si, So) :- random(X), X * N < K, !, J is K-1, M is N-1, randset(J, M, [N|Si], So). randset(K, N, Si, So) :- M is N-1, randset(K, M, Si, So). %% randseq(+K:int, +N:int, -S:list(int)) is det. % % S is a list of K integers in the range 1..N. The order is % random. % % @see randset/3. randseq(K, N, S) :- randseq(K, N, L, []), keysort(L, R), pairs_values(R, S). randseq(0, _, S, S) :- !. randseq(K, N, [Y-N|Si], So) :- random(X), X * N < K, !, random(Y), J is K-1, M is N-1, randseq(J, M, Si, So). randseq(K, N, Si, So) :- M is N-1, randseq(K, M, Si, So).