Horton

Re: [PATCH] random_normal function

Paul Ramsey <pramsey@cleverelephant.ca>

View thread
From: Paul Ramsey <pramsey@cleverelephant.ca>
To: Pgsql Hackers <pgsql-hackers@postgresql.org>
Date: 2022-12-13T23:51:11Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Round off random_normal() test results one more decimal place.

  2. Remove pg_regress' never-documented "ignore" feature.

  3. Upgrade the random.sql regression test.

  4. Invent random_normal() to provide normally-distributed random numbers.

Attachments


> On Dec 9, 2022, at 3:20 PM, Paul Ramsey <pramsey@cleverelephant.ca> wrote:
> 
> 
> 
>> On Dec 9, 2022, at 9:17 AM, Paul Ramsey <pramsey@cleverelephant.ca> wrote:
>> 
>> 
>>> On Dec 8, 2022, at 8:29 PM, Michael Paquier <michael@paquier.xyz> wrote:
>>> 
>>> On Thu, Dec 08, 2022 at 04:44:56PM -0800, Paul Ramsey wrote:
>>>> Final tme, with fixes from cirrusci.
>>> 
>>> Well, why not.  Seems like you would use that a lot with PostGIS.
>>> 
>>> #include <math.h>              /* for ldexp() */
>>> +#include <float.h>             /* for DBL_EPSILON */
>>> And be careful with the order here.
>> 
>> Should be ... alphabetical?
>> 
>>> +static void
>>> +drandom_check_default_seed()
>>> We always use (void) rather than empty parenthesis sets.
>> 
>> OK
>> 
>>> I would not leave that unchecked, so I think that you should add
>>> something in ramdom.sql.  Or would you prefer switching some of
>>> the regression tests be switched so as they use the new normal
>>> function?
>> 
>> Reading through those tests... seems like they will (rarely) fail. Is that... OK? 
>> The tests seem to be mostly worried that random() starts returning constants, which seems like a good thing to test for (is the random number generating returning randomness).
>> An obvious test for this function is that the mean and stddev converge on the supplied parameters, given enough inputs, which is actually kind of the opposite test. I use the same random number generator as the uniform distribution, so that aspect is already covered by the existing tests.
>> 
>>> (Ahem.  Bonus points for a random_string() returning a bytea, based on
>>> pg_strong_random().)
>> 
>> Would love to. Separate patch of bundled into this one?
> 
> Here's the original with suggestions applied and a random_string that applies on top of it.
> 
> Thanks!
> 
> P

Clearing up one CI failure.