Re: double precision[] storage space questions

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: greg <gregory(dot)jevardat(at)unige(dot)ch>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: double precision[] storage space questions
Date: 2015-06-12 14:18:47
Message-ID: 12042.1434118727@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

greg <gregory(dot)jevardat(at)unige(dot)ch> writes:
> I cannot find any documentation on the space taken by a double precision
> array. And the few tests I did surprise me.

> Here are a few tries I did to understand
> select pg_column_size(1.1::double precision) return 8 --- as
> expected
> select pg_column_size('{}'::double precision[]) return 16 --- ok
> maybe an array header
> select pg_column_size('{1.111}'::double precision[]) return 32 --- I
> expected 16+ sizeof(double) = 24

'{}' is a zero-dimensional array so it doesn't have the same
dimensionality information that your third case does. See
the comments at the head of
http://git.postgresql.org/gitweb/?p=postgresql.git;a=blob;f=src/include/utils/array.h;hb=HEAD

> The whole point is that in the application I work on, we store double arrays
> as bytea (using some serialization before storing the data).

TBH, that seems like a pretty silly decision. It guarantees that you
cannot do any useful manipulations of the array on the database side.

regards, tom lane

In response to

Browse pgsql-general by date

  From Date Subject
Next Message mephysto 2015-06-12 14:58:02 PostgreSQL and iptables
Previous Message greg 2015-06-12 14:08:51 double precision[] storage space questions