Re: [PATCH] random_normal function

On Tue, Dec 13, 2022 at 03:51:11PM -0800, Paul Ramsey wrote:
> Clearing up one CI failure.

+-- normal values converge on stddev == 2.0
+SELECT round(stddev(random_normal(2, 2)))
+ FROM generate_series(1, 10000);

I am not sure that it is a good idea to make a test based on a random
behavior that should tend to a normalized value. This is costly in
cycles, requiring a lot of work just for generate_series(). You could
do the same kind of thing as random() a few lines above?

+SELECT bool_and(random_string(16) != random_string(16)) AS same
+ FROM generate_series(1,8);
That should be fine in terms of impossible chances :)

+ ereport(ERROR,
+ (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
+ errmsg("return size must be non-negative")))
This could have a test, same for 0.

+#ifndef M_PI
+#define M_PI 3.14159265358979323846
+#endif
Postgres' float.h includes one version of that.
--
Michael