Horton

Re: multivariate statistics (v25)

David Rowley <david.rowley@2ndquadrant.com>

View thread View on postgr.es
From: David Rowley <david.rowley@2ndquadrant.com>
To: David Fetter <david@fetter.org>
Cc: Alvaro Herrera <alvherre@2ndquadrant.com>, Tomas Vondra <tomas.vondra@2ndquadrant.com>, Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>, dean.a.rasheed@gmail.com, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2017-03-15T02:05:38Z
Lists: pgsql-hackers 
On 15 March 2017 at 12:18, David Fetter <david@fetter.org> wrote:

>
> Is the plan to convert completely from "multivariate" to "extended?"
> I ask because I found a "multivariate" in there.
>

I get the idea that Tomas would like to keep the multivariate when it's
actually referencing multivariate stats. The idea of the rename was to
allow future expansion of the code to perhaps allow creation of stats on
expressions, which is not multivariate. If you've found multivariate
reference in an area that should be generic to extended statistics then
that's a bug and should be fixed.

I found a few of these and listed them during my review.

-- 
 David Rowley                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services

Commits

  1. Collect and use multi-column dependency stats

  2. Implement SortSupport for macaddr data type

  3. Implement multivariate n-distinct coefficients

  4. Generate fmgr prototypes automatically