Did you know ... Search Documentation:
random.pl -- Random numbers
PublicShow source

This library is derived from the DEC10 library random. Later, the core random generator was moved to C. The current version uses the SWI-Prolog arithmetic functions to realise this library. These functions are based on the GMP library.

author
- R.A. O'Keefe, V.S. Costa, L. Damas, Jan Wielemaker
See also
- Built-in function random/1: A is random(10)
Source random(-R:float) is det
Binds R to a new random float in the open interval (0.0,1.0).
See also
- setrand/1, getrand/1 may be used to fetch/set the state.
- In SWI-Prolog, random/1 is implemented by the function random_float/0.
Source random_between(+L:int, +U:int, -R:int) is semidet
Binds R to a random integer in [L,U] (i.e., including both L and U). Fails silently if U<L.
Source random(+L:int, +U:int, -R:int) is det
random(+L:float, +U:float, -R:float) is det
Generate a random integer or float in a range. If L and U are both integers, R is a random integer in the half open interval [L,U). If L and U are both floats, R is a float in the open interval (L,U).
deprecated
- Please use random/1 for generating a random float and random_between/3 for generating a random integer. Note that random_between/3 includes the upper bound, while this predicate excludes it.
Source setrand(+State) is det
Source getrand(-State) is det
Query/set the state of the random generator. This is intended for restarting the generator at a known state only. The predicate setrand/1 accepts an opaque term returned by getrand/1. This term may be asserted, written and read. The application may not make other assumptions about this term.

For compatibility reasons with older versions of this library, setrand/1 also accepts a term rand(A,B,C), where A, B and C are integers in the range 1..30,000. This argument is used to seed the random generator. Deprecated.

Errors
- existence_error(random_state, _) is raised if the underlying infrastructure cannot fetch the random state. This is currently the case if SWI-Prolog is not compiled with the GMP library.
See also
- set_random/1 and random_property/1 provide the SWI-Prolog native implementation.
Source maybe is semidet
Succeed/fail with equal probability (variant of maybe/1).
Source maybe(+P) is semidet
Succeed with probability P, fail with probability 1-P
Source maybe(+K, +N) is semidet
Succeed with probability K/N (variant of maybe/1)
Source random_perm2(?A, ?B, ?X, ?Y) is semidet
Does X=A,Y=B or X=B,Y=A with equal probability.
Source random_member(-X, +List:list) is semidet
X is a random member of List. Equivalent to random_between(1, |List|), followed by nth1/3. Fails of List is the empty list.
Compatibility
- Quintus and SICStus libraries.
Source random_select(-X, +List, -Rest) is semidet
random_select(+X, -List, +Rest) is det
Randomly select or insert an element. Either List or Rest must be a list. Fails if List is the empty list.
Compatibility
- Quintus and SICStus libraries.
Source random_subseq(+List, -Subseq, -Complement) is det
random_subseq(-List, +Subseq, +Complement) is semidet
Selects a random subsequence Subseq of List, with Complement containing all elements of List that were not selected. Each element of List is included with equal probability in either Subseq or Complement.

random_subseq/3 may also be called with Subseq and Complement bound and List unbound, which will recreate List by randomly interleaving Subseq and Complement. This mode may fail randomly, matching SICStus behavior. The failure probability corresponds to the probability of the "forward" mode selecting a Subseq/Complement combination with different lengths.

Compatibility
- SICStus 4
Source randset(+K:int, +N:int, -S:list(int)) is det
S is a sorted list of K unique random integers in the range 1..N. The implementation uses different techniques depending on the ratio K/N. For small K/N it generates a set of K random numbers, removes the duplicates and adds more numbers until |S| is K. For a large K/N it enumerates 1..N and decides randomly to include the number or not. For example:
?- randset(5, 5, S).
S = [1, 2, 3, 4, 5].          (always)
?- randset(5, 20, S).
S = [2, 7, 10, 19, 20].
See also
- randseq/3.
Source randseq(+K:int, +N:int, -List:list(int)) is det
S is a list of K unique random integers in the range 1..N. The order is random. Defined as
randseq(K, N, List) :-
      randset(K, N, Set),
      random_permutation(Set, List).
See also
- randset/3.
Source random_permutation(+List, -Permutation) is det
random_permutation(-List, +Permutation) is det
Permutation is a random permutation of List. This is intended to process the elements of List in random order. The predicate is symmetric.
Errors
- instantiation_error, type_error(list, _).
Source random_numlist(+P, +L, +U, -List) is det
Unify List with an ascending list of integers between L and U (inclusive). Each integer in the range L..U is included with probability P.
Compatibility
- SICStus 4
Source partial_list(@Term) is semidet[private]
True if Term is a partial list.

Re-exported predicates

The following predicates are exported from this file while their implementation is defined in imported modules or non-module files loaded by this module.

Source setrand(+State) is det
Source getrand(-State) is det
Query/set the state of the random generator. This is intended for restarting the generator at a known state only. The predicate setrand/1 accepts an opaque term returned by getrand/1. This term may be asserted, written and read. The application may not make other assumptions about this term.

For compatibility reasons with older versions of this library, setrand/1 also accepts a term rand(A,B,C), where A, B and C are integers in the range 1..30,000. This argument is used to seed the random generator. Deprecated.

Errors
- existence_error(random_state, _) is raised if the underlying infrastructure cannot fetch the random state. This is currently the case if SWI-Prolog is not compiled with the GMP library.
See also
- set_random/1 and random_property/1 provide the SWI-Prolog native implementation.