Skip to content

Commit 38d8176

Browse files
tglsfdctglsfdc
committed
Invent random_normal() to provide normally-distributed random numbers.
There is already a version of this in contrib/tablefunc, but it seems sufficiently widely useful to justify having it in core. Paul Ramsey Discussion: https://postgr.es/m/CACowWR0DqHAvOKUCNxTrASFkWsDLqKMd6WiXvVvaWg4pV1BMnQ@mail.gmail.com
1 parent 2673ebf commit 38d8176

File tree

10 files changed

+164
-29
lines changed

10 files changed

+164
-29
lines changed

‎doc/src/sgml/func.sgml

Lines changed: 25 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1815,6 +1815,28 @@ repeat('Pg', 4) <returnvalue>PgPgPgPg</returnvalue>
18151815
</para></entry>
18161816
</row>
18171817

1818+
<row>
1819+
<entry role="func_table_entry"><para role="func_signature">
1820+
<indexterm>
1821+
<primary>random_normal</primary>
1822+
</indexterm>
1823+
1824+
<function>random_normal</function> (
1825+
<optional> <parameter>mean</parameter> <type>double precision</type>
1826+
<optional>, <parameter>stddev</parameter> <type>double precision</type> </optional></optional> )
1827+
<returnvalue>double precision</returnvalue>
1828+
</para>
1829+
<para>
1830+
Returns a random value from the normal distribution with the given
1831+
parameters; <parameter>mean</parameter> defaults to 0.0
1832+
and <parameter>stddev</parameter> defaults to 1.0
1833+
</para>
1834+
<para>
1835+
<literal>random_normal(0.0, 1.0)</literal>
1836+
<returnvalue>0.051285419</returnvalue>
1837+
</para></entry>
1838+
</row>
1839+
18181840
<row>
18191841
<entry role="func_table_entry"><para role="func_signature">
18201842
<indexterm>
@@ -1824,7 +1846,8 @@ repeat('Pg', 4) <returnvalue>PgPgPgPg</returnvalue>
18241846
<returnvalue>void</returnvalue>
18251847
</para>
18261848
<para>
1827-
Sets the seed for subsequent <literal>random()</literal> calls;
1849+
Sets the seed for subsequent <literal>random()</literal> and
1850+
<literal>random_normal()</literal> calls;
18281851
argument must be between -1.0 and 1.0, inclusive
18291852
</para>
18301853
<para>
@@ -1848,6 +1871,7 @@ repeat('Pg', 4) <returnvalue>PgPgPgPg</returnvalue>
18481871
Without any prior <function>setseed()</function> call in the same
18491872
session, the first <function>random()</function> call obtains a seed
18501873
from a platform-dependent source of random bits.
1874+
These remarks hold equally for <function>random_normal()</function>.
18511875
</para>
18521876

18531877
<para>

‎src/backend/catalog/system_functions.sql

Lines changed: 7 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -66,6 +66,13 @@ CREATE OR REPLACE FUNCTION bit_length(text)
6666
IMMUTABLE PARALLEL SAFE STRICT COST 1
6767
RETURN octet_length($1) * 8;
6868

69+
CREATE OR REPLACE FUNCTION
70+
random_normal(mean float8 DEFAULT 0, stddev float8 DEFAULT 1)
71+
RETURNS float8
72+
LANGUAGE internal
73+
VOLATILE PARALLEL RESTRICTED STRICT COST 1
74+
AS 'drandom_normal';
75+
6976
CREATE OR REPLACE FUNCTION log(numeric)
7077
RETURNS numeric
7178
LANGUAGE sql

‎src/backend/utils/adt/float.c

Lines changed: 35 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -2743,13 +2743,11 @@ datanh(PG_FUNCTION_ARGS)
27432743

27442744

27452745
/*
2746-
* drandom - returns a random number
2746+
* initialize_drandom_seed - initialize drandom_seed if not yet done
27472747
*/
2748-
Datum
2749-
drandom(PG_FUNCTION_ARGS)
2748+
static void
2749+
initialize_drandom_seed(void)
27502750
{
2751-
float8 result;
2752-
27532751
/* Initialize random seed, if not done yet in this process */
27542752
if (unlikely(!drandom_seed_set))
27552753
{
@@ -2769,13 +2767,45 @@ drandom(PG_FUNCTION_ARGS)
27692767
}
27702768
drandom_seed_set = true;
27712769
}
2770+
}
2771+
2772+
/*
2773+
* drandom - returns a random number
2774+
*/
2775+
Datum
2776+
drandom(PG_FUNCTION_ARGS)
2777+
{
2778+
float8 result;
2779+
2780+
initialize_drandom_seed();
27722781

27732782
/* pg_prng_double produces desired result range [0.0 - 1.0) */
27742783
result = pg_prng_double(&drandom_seed);
27752784

27762785
PG_RETURN_FLOAT8(result);
27772786
}
27782787

2788+
/*
2789+
* drandom_normal - returns a random number from a normal distribution
2790+
*/
2791+
Datum
2792+
drandom_normal(PG_FUNCTION_ARGS)
2793+
{
2794+
float8 mean = PG_GETARG_FLOAT8(0);
2795+
float8 stddev = PG_GETARG_FLOAT8(1);
2796+
float8 result,
2797+
z;
2798+
2799+
initialize_drandom_seed();
2800+
2801+
/* Get random value from standard normal(mean = 0.0, stddev = 1.0) */
2802+
z = pg_prng_double_normal(&drandom_seed);
2803+
/* Transform the normal standard variable (z) */
2804+
/* using the target normal distribution parameters */
2805+
result = (stddev * z) + mean;
2806+
2807+
PG_RETURN_FLOAT8(result);
2808+
}
27792809

27802810
/*
27812811
* setseed - set seed for the random number generator

‎src/bin/pgbench/pgbench.c

Lines changed: 3 additions & 21 deletions
Original file line numberDiff line numberDiff line change
@@ -1136,8 +1136,8 @@ getGaussianRand(pg_prng_state *state, int64 min, int64 max,
11361136
Assert(parameter >= MIN_GAUSSIAN_PARAM);
11371137

11381138
/*
1139-
* Get user specified random number from this loop, with -parameter <
1140-
* stdev <= parameter
1139+
* Get normally-distributed random number in the range -parameter <= stdev
1140+
* < parameter.
11411141
*
11421142
* This loop is executed until the number is in the expected range.
11431143
*
@@ -1149,25 +1149,7 @@ getGaussianRand(pg_prng_state *state, int64 min, int64 max,
11491149
*/
11501150
do
11511151
{
1152-
/*
1153-
* pg_prng_double generates [0, 1), but for the basic version of the
1154-
* Box-Muller transform the two uniformly distributed random numbers
1155-
* are expected to be in (0, 1] (see
1156-
* https://en.wikipedia.org/wiki/Box-Muller_transform)
1157-
*/
1158-
double rand1 = 1.0 - pg_prng_double(state);
1159-
double rand2 = 1.0 - pg_prng_double(state);
1160-
1161-
/* Box-Muller basic form transform */
1162-
double var_sqrt = sqrt(-2.0 * log(rand1));
1163-
1164-
stdev = var_sqrt * sin(2.0 * M_PI * rand2);
1165-
1166-
/*
1167-
* we may try with cos, but there may be a bias induced if the
1168-
* previous value fails the test. To be on the safe side, let us try
1169-
* over.
1170-
*/
1152+
stdev = pg_prng_double_normal(state);
11711153
}
11721154
while (stdev < -parameter || stdev >= parameter);
11731155

‎src/common/pg_prng.c

Lines changed: 36 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -19,11 +19,17 @@
1919

2020
#include "c.h"
2121

22-
#include <math.h> /* for ldexp() */
22+
#include <math.h>
2323

2424
#include "common/pg_prng.h"
2525
#include "port/pg_bitutils.h"
2626

27+
/* X/Open (XSI) requires <math.h> to provide M_PI, but core POSIX does not */
28+
#ifndef M_PI
29+
#define M_PI 3.14159265358979323846
30+
#endif
31+
32+
2733
/* process-wide state vector */
2834
pg_prng_state pg_global_prng_state;
2935

@@ -235,6 +241,35 @@ pg_prng_double(pg_prng_state *state)
235241
return ldexp((double) (v >> (64 - 52)), -52);
236242
}
237243

244+
/*
245+
* Select a random double from the normal distribution with
246+
* mean = 0.0 and stddev = 1.0.
247+
*
248+
* To get a result from a different normal distribution use
249+
* STDDEV * pg_prng_double_normal() + MEAN
250+
*
251+
* Uses https://en.wikipedia.org/wiki/Box%E2%80%93Muller_transform
252+
*/
253+
double
254+
pg_prng_double_normal(pg_prng_state *state)
255+
{
256+
double u1,
257+
u2,
258+
z0;
259+
260+
/*
261+
* pg_prng_double generates [0, 1), but for the basic version of the
262+
* Box-Muller transform the two uniformly distributed random numbers are
263+
* expected to be in (0, 1]; in particular we'd better not compute log(0).
264+
*/
265+
u1 = 1.0 - pg_prng_double(state);
266+
u2 = 1.0 - pg_prng_double(state);
267+
268+
/* Apply Box-Muller transform to get one normal-valued output */
269+
z0 = sqrt(-2.0 * log(u1)) * sin(2.0 * M_PI * u2);
270+
return z0;
271+
}
272+
238273
/*
239274
* Select a random boolean value.
240275
*/

‎src/include/catalog/catversion.h

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -57,6 +57,6 @@
5757
*/
5858

5959
/* yyyymmddN */
60-
#define CATALOG_VERSION_NO 202301091
60+
#define CATALOG_VERSION_NO 202301092
6161

6262
#endif

‎src/include/catalog/pg_proc.dat

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3359,6 +3359,10 @@
33593359
{ oid => '1598', descr => 'random value',
33603360
proname => 'random', provolatile => 'v', proparallel => 'r',
33613361
prorettype => 'float8', proargtypes => '', prosrc => 'drandom' },
3362+
{ oid => '8074', descr => 'random value from normal distribution',
3363+
proname => 'random_normal', provolatile => 'v', proparallel => 'r',
3364+
prorettype => 'float8', proargtypes => 'float8 float8',
3365+
prosrc => 'drandom_normal' },
33623366
{ oid => '1599', descr => 'set random seed',
33633367
proname => 'setseed', provolatile => 'v', proparallel => 'r',
33643368
prorettype => 'void', proargtypes => 'float8', prosrc => 'setseed' },

‎src/include/common/pg_prng.h

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -55,6 +55,7 @@ extern uint32 pg_prng_uint32(pg_prng_state *state);
5555
extern int32 pg_prng_int32(pg_prng_state *state);
5656
extern int32 pg_prng_int32p(pg_prng_state *state);
5757
extern double pg_prng_double(pg_prng_state *state);
58+
extern double pg_prng_double_normal(pg_prng_state *state);
5859
extern bool pg_prng_bool(pg_prng_state *state);
5960

6061
#endif /* PG_PRNG_H */

‎src/test/regress/expected/random.out

Lines changed: 28 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -51,3 +51,31 @@ SELECT AVG(random) FROM RANDOM_TBL
5151
-----
5252
(0 rows)
5353

54+
-- now test random_normal()
55+
TRUNCATE random_tbl;
56+
INSERT INTO random_tbl (random)
57+
SELECT count(*)
58+
FROM onek WHERE random_normal(0, 1) < 0;
59+
INSERT INTO random_tbl (random)
60+
SELECT count(*)
61+
FROM onek WHERE random_normal(0) < 0;
62+
INSERT INTO random_tbl (random)
63+
SELECT count(*)
64+
FROM onek WHERE random_normal() < 0;
65+
INSERT INTO random_tbl (random)
66+
SELECT count(*)
67+
FROM onek WHERE random_normal(stddev => 1, mean => 0) < 0;
68+
-- expect similar, but not identical values
69+
SELECT random, count(random) FROM random_tbl
70+
GROUP BY random HAVING count(random) > 3;
71+
random | count
72+
--------+-------
73+
(0 rows)
74+
75+
-- approximately check expected distribution
76+
SELECT AVG(random) FROM random_tbl
77+
HAVING AVG(random) NOT BETWEEN 400 AND 600;
78+
avg
79+
-----
80+
(0 rows)
81+

‎src/test/regress/sql/random.sql

Lines changed: 24 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -42,3 +42,27 @@ SELECT random, count(random) FROM RANDOM_TBL
4242

4343
SELECT AVG(random) FROM RANDOM_TBL
4444
HAVING AVG(random) NOT BETWEEN 80 AND 120;
45+
46+
-- now test random_normal()
47+
48+
TRUNCATE random_tbl;
49+
INSERT INTO random_tbl (random)
50+
SELECT count(*)
51+
FROM onek WHERE random_normal(0, 1) < 0;
52+
INSERT INTO random_tbl (random)
53+
SELECT count(*)
54+
FROM onek WHERE random_normal(0) < 0;
55+
INSERT INTO random_tbl (random)
56+
SELECT count(*)
57+
FROM onek WHERE random_normal() < 0;
58+
INSERT INTO random_tbl (random)
59+
SELECT count(*)
60+
FROM onek WHERE random_normal(stddev => 1, mean => 0) < 0;
61+
62+
-- expect similar, but not identical values
63+
SELECT random, count(random) FROM random_tbl
64+
GROUP BY random HAVING count(random) > 3;
65+
66+
-- approximately check expected distribution
67+
SELECT AVG(random) FROM random_tbl
68+
HAVING AVG(random) NOT BETWEEN 400 AND 600;

0 commit comments

Comments
 (0)
pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy