From: Tom Lane <tgl@sss.pgh.pa.us>
To: "David G. Johnston" <david.g.johnston@gmail.com>
cc: "robb@helical.com" <robb@helical.com>,
        "pgsql-bugs@lists.postgresql.org" <pgsql-bugs@lists.postgresql.org>
Subject: Re: BUG #18390: exponentiation produces float datatype,
 but nth-root produces integer
In-reply-to: 
 <CAKFQuwYyJT+HFcOqGA5nawKA4yOnMdbGdqRkKbBLZ5KWFzcg7Q@mail.gmail.com>
References: <18390-03499fe4f5a0ac71@postgresql.org>
 <CAKFQuwYyJT+HFcOqGA5nawKA4yOnMdbGdqRkKbBLZ5KWFzcg7Q@mail.gmail.com>
Comments: In-reply-to "David G. Johnston" <david.g.johnston@gmail.com>
	message dated "Wed, 13 Mar 2024 05:07:35 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <2043863.1710339130.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 13 Mar 2024 10:12:10 -0400
Message-ID: <2043864.1710339130@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2043864.1710339130%40sss.pgh.pa.us>
Precedence: bulk

"David G. Johnston" <david.g.johnston@gmail.com> writes:
> On Tuesday, March 12, 2024, PG Bug reporting form <noreply@postgresql.or=
g>
> wrote:
>> However, an nth-root calculation from that float result does NOT produc=
e a
>> new float result:

> The issue is you expect dividing two integers to produce a float but tha=
t
> isn=E2=80=99t how that works.  Dividing two integers produces an integer=
.  In this
> case zero, or .1 rounded down/truncated.

Possibly adding to the OP's confusion: not one of these examples
contains any float arithmetic whatsoever.  They're of type numeric.

=3D# select pg_typeof(2.5937424601000000^(1.0/10));
 pg_typeof =

-----------
 numeric
(1 row)

See
https://www.postgresql.org/docs/current/sql-syntax-lexical.html#SQL-SYNTAX=
-CONSTANTS-NUMERIC
which says

    A numeric constant that contains neither a decimal point nor an
    exponent is initially presumed to be type integer if its value
    fits in type integer (32 bits); otherwise it is presumed to be
    type bigint if its value fits in type bigint (64 bits); otherwise
    it is taken to be type numeric. Constants that contain decimal
    points and/or exponents are always initially presumed to be type
    numeric.

A cast to float would occur only if the value is fed to an operator
that doesn't come in an exactly matching datatype.

			regards, tom lane