road.thepath no longer in pg_stats?

Started by Alvaro Herreraalmost 16 years ago2 messageshackers
Jump to latest
#1Alvaro Herrera
alvherre@2ndquadrant.com

Hi,

I just noticed that the road.thepath column in the regression database
is no longer in pg_stats for some reason after ANALYZE, so the example
in section 14.2 in the docs (Statistics Used by the Planner) is now
wrong.

What I'm trying to do is replace that exceedingly wide output with
something along these lines instead:

SELECT attname, n_distinct, array_to_string(most_common_vals, E'\n') as most_common_vals
FROM pg_stats
WHERE tablename = 'road';

but this is failing because of missing values for that column, as well
as there being two rows for the column that is there (stainherint=f
producing the other one)

--
Alvaro Herrera http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support

#2Tom Lane
tgl@sss.pgh.pa.us
In reply to: Alvaro Herrera (#1)
Re: road.thepath no longer in pg_stats?

Alvaro Herrera <alvherre@commandprompt.com> writes:

I just noticed that the road.thepath column in the regression database
is no longer in pg_stats for some reason after ANALYZE, so the example
in section 14.2 in the docs (Statistics Used by the Planner) is now
wrong.

It's an old example :-( ... a quick check shows no version since 7.3
generates stats for that column.

but this is failing because of missing values for that column, as well
as there being two rows for the column that is there (stainherint=f
producing the other one)

Two rows are expected now, since that table has children --- there
should be one for inherited = t and one for inherited = f. Probably
wouldn't hurt to explain that explicitly here.

regards, tom lane