SRFI 27
Source of Random Bits
TOC »
- SRFI 27
- Documentation
- SRFI 27
- Usage
- current-random-source
- current-random-integer
- current-random-real
- registered-random-sources
- registered-random-source
- random-source?
- make-random-source
- new-random-source
- random-source-name
- random-source-documentation
- random-source-log2-period
- random-source-maximum-range
- random-source-entropy-source
- random-source-entropy-source-set!
- random-source-state-ref
- random-source-state-set!
- random-source-randomize!
- random-source-pseudo-randomize!
- random-source-make-integers
- random-source-make-reals
- random-source-make-u8vectors
- random-source-make-f64vectors
- current-entropy-source
- registered-entropy-sources
- registered-entropy-source
- entropy-source?
- make-entropy-source
- new-entropy-source
- entropy-source-name
- entropy-source-documentation
- entropy-source-u8
- entropy-source-f64
- entropy-source-u8vector
- entropy-source-f64vector
- entropy-source-integer
- Random Sources
- Entropy Sources
- Procedure Entropy Source
- Port Entropy Source
- Simple Distributions
- Vector Distributions
- Other Distributions
- Usage
- make-random-normals
- make-random-exponentials
- make-random-triangles
- make-random-poissons
- make-random-bernoullis
- make-random-binomials
- make-random-geometrics
- make-random-lognormals
- make-random-cauchys
- make-random-gammas
- make-random-erlangs
- make-random-paretos
- make-random-levys
- make-random-weibulls
- SRFI 27
- Examples
- Notes
- Bugs and Limitations
- Requirements
- Author
- Repository
- Version history
- License
Documentation
A Chicken implementation of SRFI 27.
SRFI 27
Most procedures are per the SRFI 27 specification document. Only the extensions are documented here.
The most obvious extensions are support for multiple random number generators and multiple entropy sources.
Usage
(import (srfi 27))
current-random-source
- current-random-source #!optional RANDOM-SOURCEprocedure
Parameter for the default random-source.
The variable default-random-source is the initial (current-random-source).
The initial random-source is an instance of 'mrg32k3a, per SRFI 27.
current-random-integer
- current-random-integerprocedure
Returns an integer random generator from the (current-random-source).
Like random-integer but w/ (current-random-source), & a procedure, not a variable.
current-random-real
- current-random-realprocedure
Returns a real random generator from the (current-random-source).
Like random-real but w/ (current-random-source), & a procedure, not a variable.
registered-random-sources
- registered-random-sourcesprocedure
Returns a list of the registered random-source names.
registered-random-source
- registered-random-source NAMEprocedure
Returns the random-source creator for the specified NAME or #f if not registered.
random-source?
- random-source? OBJprocedure
random-source predicate.
- check-random-source LOC OBJ #!optional NAMprocedure
Raise error when OBJ not a random-source. NAM is the optional parameter name.
- error-random-source LOC OBJ #!optional NAMprocedure
Raise error for OBJ not a random-source. NAM is the optional parameter name.
make-random-source
- make-random-source #!optional SOURCE current-random-sourceprocedure
SOURCE is either a random-source or the symbol of a registered random-source, the name.
new-random-source
- new-random-source #!optional RANDOM-SOURCE current-random-sourceprocedure
Same as (make-random-source RANDOM-SOURCE).
random-source-name
- random-source-name RANDOM-SOURCEprocedure
The symbolic name of the RANDOM-SOURCE.
random-source-documentation
- random-source-documentation RANDOM-SOURCEprocedure
Some more information for the RANDOM-SOURCE.
random-source-log2-period
- random-source-log2-period RANDOM-SOURCEprocedure
The period of the RANDOM-SOURCE as a power of 2. A fixnum.
random-source-maximum-range
- random-source-maximum-range RANDOM-SOURCEprocedure
The largest integer the RANDOM-SOURCE can produce without resort to a bignum representation. Currently an inexact-integer flonum, even for 64-bit CPUs.
random-source-entropy-source
- random-source-entropy-source RANDOM-SOURCEprocedure
The current entropy-source for the RANDOM-SOURCE.
random-source-entropy-source-set!
- random-source-entropy-source-set! RANDOM-SOURCE ENTROPY-SOURCEprocedure
Changes the current entropy-source for the RANDOM-SOURCE.
random-source-state-ref
- random-source-state-ref RANDOM-SOURCEprocedure
Per the SRFI 27 specification the random-state is a valid external representation of RANDOM-SOURCE state.
Note a random-source is not required to provide an external state, in which case #f is returned.
random-source-state-set!
- random-source-state-set! RANDOM-SOURCE RANDOM-STATEprocedure
Only a RANDOM-STATE produced by a random-source-state-ref of the same random-source is acceptable.
Note a random-source is not required to provide an external state, in which case no state change.
random-source-randomize!
random-source-pseudo-randomize!
random-source-make-integers
- random-source-make-integers RANDOM-SOURCEprocedure
The result is a procedure: (random-source-make-integers RANDOM-SOURCE)
random-source-make-reals
- random-source-make-reals RANDOM-SOURCE #!optional PRECISIONprocedure
The result is a procedure: (random-source-make-reals RANDOM-SOURCE PRECISION)
PRECISION maybe #f, in which case the value is flonum-epsilon.
random-source-make-u8vectors
- random-source-make-u8vectors RANDOM-SOURCEprocedure
Returns a procedure of one argument, the length of the generated vector, the returns a vector of random 8-bit unsigned values, a SRFI 4 u8vector.
The variable random-u8vector is a procedure: (random-source-make-u8vectors default-random-source)
random-source-make-f64vectors
- random-source-make-f64vectors RANDOM-SOURCE #!optional PRECISIONprocedure
Returns a procedure of one argument, the length of the generated vector, the returns a vector of random 64-bit floating-point values, a SRFI 4 f64vector.
The variable random-f64vector is a procedure: (random-source-make-f64vectors default-random-source)
current-entropy-source
- current-entropy-source #!optional ENTROPY-SOURCEprocedure
Parameter for the default entropy-source.
default-entropy-source is the initial (current-entropy-source).
The initial entropy-source is an instance of 'system-clock.
registered-entropy-sources
- registered-entropy-sourcesprocedure
Returns a list of the registered entropy-source names.
registered-entropy-source
- registered-entropy-source NAMEprocedure
Returns the entropy-source creator for the specified NAME or #f if not registered.
entropy-source?
- entropy-source? OBJprocedure
entropy-source predicate.
- check-entropy-source LOC OBJ #!optional NAMprocedure
Raise error when OBJ not an entropy-source. NAM is the optional parameter name.
- error-entropy-source LOC OBJ #!optional NAMprocedure
Raise error for OBJ not an entropy-source. NAM is the optional parameter name.
make-entropy-source
- make-entropy-source #!optional SOURCE current-entropy-sourceprocedure
SOURCE is either a entropy-source or the symbol of a registered entropy-source, the name.
new-entropy-source
- new-entropy-source ENTROPY-SOURCEprocedure
Same as (make-random-entropy ENTROPY-SOURCE).
entropy-source-name
- entropy-source-name ENTROPY-SOURCEprocedure
The symbolic name of the ENTROPY-SOURCE.
entropy-source-documentation
- entropy-source-documentation ENTROPY-SOURCEprocedure
Some more information for the ENTROPY-SOURCE.
entropy-source-u8
- entropy-source-u8 ENTROPY-SOURCEprocedure
Returns a non-negative fixnum from the ENTROPY-SOURCE.
entropy-source-f64
- entropy-source-f64 ENTROPY-SOURCEprocedure
Returns a flonum from the ENTROPY-SOURCE.
entropy-source-u8vector
- entropy-source-u8vector ENTROPY-SOURCE LENGTH #!optional U8VECTORprocedure
Returns a u8vector with the 0 thru LENGTH-1 elements filled with non-negative fixnums from the ENTROPY-SOURCE. If U8VECTOR is supplied then this is used otherwise a new vector is returned.
entropy-source-f64vector
- entropy-source-f64vector ENTROPY-SOURCE LENGTH #!optional F64VECTORprocedure
Returns a f64vector with the 0 thru LENGTH-1 elements filled with flonums from the ENTROPY-SOURCE. If F64VECTOR is supplied then this is used otherwise a new vector is returned.
entropy-source-integer
- entropy-source-integer ENTROPY-SOURCEprocedure
Returns an integer from the ENTROPY-SOURCE.
Random Sources
A random source module must be loaded before it can be used. This maybe obvious but a call of (make-random-source 'moa) will fail unless the extension "moa" is loaded.
See make-random-source for use of the NAME.
System random number generator
Usage
(import random-system)
make-random-source-system
- make-random-source-systemprocedure
Registered NAME is 'system.
Pierre L'Ecuyer's Multiple Recursive Generator 32k3a random number generator
Usage
(import mrg32k3a)
make-random-source-mrg32k3a
- make-random-source-mrg32k3aprocedure
Registered NAME is 'mrg32k3a.
George Marsaglia's Multiply With Carry random number generator
Usage
(import mwc)
make-random-source-mwc
- make-random-source-mwcprocedure
Registered NAME is 'mwc.
George Marsaglia's Mother Of All random number generator
Usage
(import moa)
make-random-source-moa
- make-random-source-moaprocedure
Registered NAME is 'moa.
Entropy Sources
An entropy source module must be loaded before it can be used.
System
Usage
(import entropy-system)
make-entropy-source-system
- make-entropy-source-systemprocedure
Registered NAME is 'system,
System Clock
Usage
(import entropy-clock)
make-entropy-source-system-clock
- make-entropy-source-system-clockprocedure
Registered NAME is 'system-clock.
Windows Crypt Device
Usage
(import entropy-windows)
make-entropy-source-crypt
- make-entropy-source-crypt #!optional BUFFER-LENGTHprocedure
Registered NAME is 'crypt.
Unix Random Device
Usage
(import entropy-unix)
make-entropy-source-urandom-device
- make-entropy-source-urandom-deviceprocedure
Registered NAME is 'random-device.
make-entropy-source-random-device
- make-entropy-source-random-deviceprocedure
Registered NAME is 'urandom-device.
Procedure Entropy Source
U8PROC is a procedure of no arguments returning a fixnum in the range 0 .. 255.
F64PROC is a procedure of no arguments returning a finite flonum.
Usage
(import entropy-procedure)
make-entropy-source/procedures
- (make-entropy-source/procedures U8PROC F64PROC [name: (NAME (gensym 'procedures-))] [docu: (DOCU "Entropy from procedures")]) -> entropy-sourceprocedure
Returns an unregistered entropy-source built from the supplied U8PROC and F64PROC procedures.
make-entropy-source/f64procedure
- (make-entropy-source/f64procedure F64PROC [name: (NAME (gensym 'procedures-))] [docu: (DOCU "Entropy from procedures")]) -> entropy-sourceprocedure
Returns an unregistered entropy-source built from the supplied F64PROC procedure.
Port Entropy Source
Usage
(import entropy-port)
entropy-port-lifetime
- entropy-port-lifetime #!optional SECONDSparameter
The number of SECONDS an entropy port is kept open without any activity.
SECONDS is a positive real or a boolean: #t, to reset the default seconds, or #f, to determine the lifetime by GC finalization.
make-entropy-source/port
- (make-entropy-source/port PORT [name: (NAME (gensym 'port-))] [docu: (DOCU "Entropy from an open port")]) -> entropy-sourceprocedure
Returns an unregistered entropy-source built from the supplied PORT, which is treated as a binary stream.
The PORT is kept open and must be closed by the caller, if at all.
make-entropy-source/port-open
- (make-entropy-source/port-open OPENER [name: (NAME (gensym 'port-))] [docu: (DOCU "Entropy from port")]) -> entropy-sourceprocedure
Returns an unregistered entropy-source built from the supplied OPENER.
- OPENER
- (-> port) ; returns an opened port.
The returned port has an (entropy-port-lifetime) so the port may be closed and OPENER called more than once.
make-entropy-source/port-open-timed
- (make-entropy-source/port-open-timed OPENER SECONDS [name: (NAME (gensym 'timed-port-))] [docu: (DOCU "Entropy from timed open port")]) -> entropy-sourceprocedure
Returns an unregistered entropy-source built from the supplied OPENER & SECONDS.
- SECONDS
- real ; timeout.
- OPENER
- (-> port) ; interpreted as with make-entropy-source/port-open.
make-entropy-source/file
- (make-entropy-source/file NAMSTR [name: (NAME (gensym 'file-))] [docu: (DOCU (string-append "Entropy from file \"" namstr "\""))]) -> entropy-sourceprocedure
Returns an unregistered entropy-source using the file named by the pathname NAMSTR.
The opened port for the file NAMSTR is interpreted as with make-entropy-source/port-open.
make-entropy-source/file-timed
- (make-entropy-source/file-timed NAMSTR SECONDS [name: (NAME (gensym 'file-))] [docu: (DOCU (string-append "Entropy from file \"" namstr "\""))]) -> entropy-sourceprocedure
Returns an unregistered entropy-source using the file named by the pathname NAMSTR & a lifetime restricted to SECONDS.
- SECONDS
- real ; timeout.
The opened port for the file NAMSTR is interpreted as with make-entropy-source/port-open.
Simple Distributions
Provides exact and inexact generators. All return 2 values: the generator itself and a procedure returning the argument values used to parameterize the generator. The second return value can be ignored.
Usage
(import (srfi 27 uniform-random))
make-uniform-random-integers
- (make-uniform-random-integers [high: (HIGH #f)] [low: (LOW 0)] [precision: (PRECISION 1)] [source: (SOURCE (current-random-source))]) -> (-> integer)procedure
LOW, HIGH and PRECISION are integers.
SOURCE is a random-source.
HIGH, if not supplied, is the (- (random-source-maximum-range source) 1).
The generator returns integers in the range LOW .. HIGH with an index of PRECISION. Unlike a (random-integer) result negative integers are possible.
(current-random-integer) <> (make-uniform-random-integers). (current-random-integer) returns ((integer) -> integer). (make-uniform-random-integers) returns (() -> integer).
make-uniform-random-reals
- (make-uniform-random-reals [precision: PRECISION] [source: (SOURCE (current-random-source))]) -> (-> real)procedure
The generator is as (random-source-make-reals SOURCE PRECISION).
PRECISION maybe #f, in which case the value is flonum-epsilon.
SOURCE is a random-source.
(current-random-real) = (make-uniform-random-reals). Both return procedures of (() -> real).
Vector Distributions
The procedures named as per make-random-... return a procedure of one argument N, the length of the vector to be created.
The nomenclature vector% is used to indicate a disjoint type union of vector, f32vector, f64vector. If the user is really keen to access the API for this type it is available in the, undocumented, "srfi-27-vector-support" module.
Usage
(import (srfi 27 vector))
make-random-permutations
random-permutation!
- (make-random-permutations [randoms: (RANDOMS current-random-integer)]) -> (integer -> vector)procedure
- (random-permutation! VECTOR [randoms: (RANDOMS current-random-integer)]) -> vectorprocedure
Performs the "Knuth shuffle" (or "Fisher-Yates shuffle").
Fills VECTOR with a random permutation of the finite set {0 ... N-1}, where N = (vector-length VECTOR).
RANDOMS is a procedure: (exact-real) -> exact-real
(From Gambit)
make-random-vector
random-vector!
- (make-random-vector [randoms: (RANDOMS current-random-real)]) -> (integer -> vector)procedure
- (random-vector! VECTOR% [randoms: (RANDOMS current-random-real)]) -> vector%procedure
Fills VECTOR with inexact real random numbers from the random distribution generator randoms.
RANDOMS is a procedure: () -> inexact-real
make-random-hollow-sphere
random-hollow-sphere!
- (make-random-hollow-sphere [mu: (MU 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> (integer -> vector)procedure
- (random-hollow-sphere! VECTOR% [mu: (MU 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> vector%procedure
Fills VECTOR% with inexact real random numbers the sum of whose squares are equal to 1.
Thinking of VECTOR% as coordinates in space of dimension N = (vector%-length VECTOR%), the coordinates are uniformly distributed over the surface of the unit n-sphere.
(From Gambit)
make-random-solid-sphere
random-solid-sphere!
- (make-random-solid-sphere [mu: (MU 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> (integer -> vector)procedure
- (random-solid-sphere! VECTOR% [mu: (MU 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> vector%procedure
Fills VECTOR% with inexact real random numbers the sum of whose squares are less than 1.
Thinking of VECTOR% as coordinates in space of dimension N = (vector%-length VECTOR%), the coordinates are uniformly distributed within the unit n-sphere.
(From Gambit)
Other Distributions
Provides generators for some common distributions ranging over some subset of the inexact-reals. The domain of the generators depends, of course, on the specific distribution.
All return 2 values: the generator itself and a procedure returning the argument values used to parameterize the generator. The second return value can be ignored.
The identifiers matching *make-random-<distribution> are reserved.
Usage
(import (srfi 27 distributions))
All distributions are available from the omnibus module (srfi 27 distributions), and individually as the (srfi 27 <distribution>) module.
make-random-normals
- (make-random-normals [mu: (MU 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
MU is real.
SIGMA is nonzero-real.
RANDOMS is a (-> real)
(import (srfi 27 normals))
make-random-exponentials
- (make-random-exponentials [mu: (MU 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
MU is real in [0 1].
RANDOMS is a (-> real)
(import (srfi 27 exponentials))
make-random-triangles
- (make-random-triangles [s: (S 0)] [m: (M 1/2)] [l: (L 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
S is real.
M is real in [S L].
L is real in ]S +inf.0[.
RANDOMS is a (-> real)
(import (srfi 27 triangles))
make-random-poissons
- (make-random-poissons [mu: (MU 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> integer)procedure
MU is nonnegative-real.
RANDOMS is a (-> real)
(import (srfi 27 poissons))
make-random-bernoullis
- (make-random-bernoullis [p: (P 1/2)] [randoms: (RANDOMS (current-random-real))]) -> (-> boolean)procedure
P is real in [0 1].
RANDOMS is a (-> real)
(import (srfi 27 bernoullis))
make-random-binomials
- (make-random-binomials [t: (T 1)] [p: (P 1/2)] [randoms: (RANDOMS (current-random-real))]) -> (-> integer)procedure
T is cardinal-integer.
P is real in [0 1].
RANDOMS is a (-> real)
(import (srfi 27 binomials))
make-random-geometrics
- (make-random-geometrics [p: (P 1/2)] [randoms: (RANDOMS (current-random-real))])) -> (-> integer)procedure
P is real in [0 1].
RANDOMS is a (-> real)
(import (srfi 27 geometrics))
make-random-lognormals
- (make-random-lognormals [mu: (MU 1)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
MU is nonzero-real.
SIGMA is nonnegative-real.
RANDOMS is a (-> real)
(import (srfi 27 lognormals))
make-random-cauchys
- (make-random-cauchys [median: (MEDIAN 0)] [sigma: (SIGMA 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
MEDIAN is real.
SIGMA is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 cauchys))
make-random-gammas
- (make-random-gammas [alpha: (ALPHA 1)] [theta: (THETA 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
ALPHA is positive-real.
THETA is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 gammas))
make-random-erlangs
- (make-random-erlangs [alpha: (ALPHA 1)] [theta: (THETA 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
ALPHA is positive-real.
THETA is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 erlangs))
make-random-paretos
- (make-random-paretos [alpha: (ALPHA 1)] [xmin: (XMIN 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
ALPHA is positive-real.
XMIN is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 paretos))
make-random-levys
- (make-random-levys [gamma: (GAMMA 1)] [delta: (DELTA 0)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
GAMMA is nonnegative-real.
DELTA is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 levys))
make-random-weibulls
- (make-random-weibulls [shape: (SHAPE 1)] [scale: (SCALE 1)] [randoms: (RANDOMS (current-random-real))]) -> (-> real)procedure
SHAPE is positive-real.
SCALE is positive-real.
RANDOMS is a (-> real)
(import (srfi 27 weibulls))
Examples
- A procedure for creating K-vector (whatever that is) generators. The use of George Marsaglia's Multiply With Carry random-source and the "/dev/random" entropy-source are for exposition only.
(import entropy-unix entropy-port mwc (srfi 27) (srfi 27 uniform-random) (srfi 27 vector)) (define (make-random-k-vector #!optional (k 1024) (rs (make-random-source 'mwc)) es) (let ( (new-es (or es ;don't keep open port around too much (parameterize ((entropy-port-lifetime 30)) (make-entropy-source-random-device)))) (old-es (random-source-entropy-source rs)) ) (dynamic-wind (lambda () (random-source-entropy-source-set! rs new-es) ) (lambda () (let* ( (rnd (make-uniform-random-integers low: (- k) high: (+ k) precision: 2 source: rs)) (vecgen (make-random-vector randoms: rnd)) ) (lambda (#!optional (k k)) (vecgen k)) ) ) (lambda () (random-source-entropy-source-set! rs old-es) ) ) ) )
- Providing a fixed entropy-source, a constant seed.
(import data-structures entropy-procedure) (define (make-entropy-constant n) (make-entropy-source/f64procedure (constantly (exact->inexact n)) name: (string->uninterned-symbol (conc "entropy-constant-" n)) docu: (conc "Entropy constant " n)) )
- A silly OO version of a random distribution generator
(import (srfi 1) (srfi 13)) (import coops) (import (srfi 27 normals) (srfi 27 exponentials)) ;; Named (has a name) "concept" (define-generic (name obj)) (define-class <named> () ( (namsym reader: name) ) ) ;; Parameterized extension "concept" (define-class <parameterized> () (ctor params src)) ;; Moves forward thru a set of values "concept" (define-generic (next-value obj)) (define-class <stepper> () (nxtval)) (define-method (next-value (obj <stepper>)) ((slot-value obj 'nxtval))) (define-generic (reset obj)) (define-method (reset (obj <stepper>)) (let-values ( ((gen _) (apply (slot-value obj 'ctor) `(,@(slot-value obj 'params) ,(slot-value obj 'src)))) ) (set! (slot-value obj 'nxtval) gen) ) ) ;; Parameterized generative set of random values "concept" (define-class <random-distribution> (<named> <parameterized> <stepper>) ( ;holds any value, temporarily tmpval ) ) ;; Create an instance of <random-distribution> where the arguments are ;; the same as the documented procedural distribution API. ;; ;; SRFI 27 API: ({some distribution constructor} arg...) ;; OO API: (make-random-distribution {some distribution constructor} arg...) (define-syntax make-random-distribution (syntax-rules () ((_ ?ctor ?arg0 ...) ;use tmpval to hold the ctor call form (make <random-distribution> 'tmpval (list ?ctor ?arg0 ...)) ) ) ) (define-method (initialize-instance (obj <random-distribution>)) ; The 'ctor' must be a globally defined procedure compiled with ; procedure-information. So if following nomenclature then the last ; procedure name element will be the kind of distribution: ; make-random-<distribution>. And the <random-source> will be the ; last argument. ; ; (I do not endorse this kind of "auto-magic". For example only.) (let* ( (tmpval (or (slot-value obj 'tmpval) `(,make-random-normals))) (ctor (car tmpval) ) (procinfo (procedure-information ctor)) (procname (and procinfo (pair? procinfo) (symbol->string (car procinfo)))) (procname (and procname (and-let* ((kndpos (string-index-right procname #\-))) (substring/shared procname (add1 kndpos))))) (procname (or procname "impossible")) (dstr-vals (receive (apply ctor (cdr tmpval)))) (params (and (<= 2 (length dstr-vals)) (receive ((second dstr-vals))))) ) ;"free" the temp slot (set! (slot-value obj 'tmpval) #f) ;initialize state (set! (slot-value obj 'namsym) (string->symbol procname)) (set! (slot-value obj 'nxtval) (first dstr-vals)) (set! (slot-value obj 'ctor) ctor) (set! (slot-value obj 'params) (and params (drop-right params 1))) (set! (slot-value obj 'src) (and params (last params))) ) ) ;; Use it (import coops-extras) (define expn-rd (make-random-distribution make-random-exponentials mu: 1/2)) (describe-object expn-rd) ;coops instance of class `<random-distribution>': ;tmpval: #f ;namsym: exponentials ; ctor: #<procedure (srfi-27-exponentials#make-random-exponentials . tmp302303)> ;params: (1/2) ; src: #<procedure> ;nxtval: #<procedure (f_1191)> (next-value expn-rd) ;=> ...
Notes
- Portions by Sebastian Egner and Brad Lucier. (See the "book" SRFI 27 implementation.)
Bugs and Limitations
- Support for 64-bit platforms is poor. Need to provide core random integer generators which are defined over the 64-bit integer range. Currently 64-bit fixnum bounds are processed as if using a 32-bit platform. However, due to the bignum implementation, results in the 64-bit fixnum range are coerced to fixnum.
- The random distribution API is poorly documented.
- The implementation API is not documented. See the source distribution.
- Use of an entropy-source as a random-source is impossible without the undocumented implementation API.
Requirements
srfi-1 vector-lib miscmacros check-errors timed-resource
Author
- Kon Lovett
- Ported to Chicken 5 by Kon Lovett and Sergey Goldgaber
Repository
This egg is hosted on the CHICKEN Subversion repository:
https://anonymous@code.call-cc.org/svn/chicken-eggs/release/5/srfi-27
If you want to check out the source code repository of this egg and you are not familiar with Subversion, see this page.
Version history
- 4.2.2
- 4.2.1
- 4.2.0
- Add entropy-system entropy-source & random-system random-source.
- 4.1.6
- Remove dependencies.
- 4.1.5
- Fix entropy-windows compilation error.
- 4.1.4
- Fix entropy-windows compilation error.
- 4.1.3
- Fix modulo of a flonum issue.
- 4.1.2
- Remove broken 32bit platform CHICKEN 4 optimization.
- 4.1.1
- Update test runner.
- 4.1.0
- Remove deprecated exports. Non-SRFI-27 module renaming to support the (srfi 27 <extension>) form.
- 4.0.0
- Ported to Chicken 5
License
Copyright (C) 2010-2022 Kon Lovett. All rights reserved.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the Software), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED ASIS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Does not supercede any restrictions found in the source code.