From: Tom Lane <tgl@sss.pgh.pa.us>
To: bingyanli@email.ncu.edu.cn
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #18951: Precision loss in inner join while using SUM
 aggregate function
In-reply-to: <18951-3092d80beab3d2ef@postgresql.org>
References: <18951-3092d80beab3d2ef@postgresql.org>
Comments: In-reply-to PG Bug reporting form <noreply@postgresql.org>
	message dated "Sun, 08 Jun 2025 14:38:55 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <183008.1749397821.1@sss.pgh.pa.us>
Date: Sun, 08 Jun 2025 11:50:21 -0400
Message-ID: <183009.1749397821@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/183009.1749397821%40sss.pgh.pa.us>
Precedence: bulk

PG Bug reporting form <noreply@postgresql.org> writes:
> -- judgement: s1 = s2 + s3, However, s2 is not correctly computed

[ shrug... ] If you are expecting exact results, don't use
floating-point arithmetic (ie, use "numeric" not "double precision").
This example, which will result in adding values of enormously
different magnitudes in various orders, is tailor-made to
exhibit roundoff error.

https://www.postgresql.org/docs/current/datatype-numeric.html#DATATYPE-FLOAT

			regards, tom lane