Re: Suitable Index for my Table
От | Kevin Grittner |
---|---|
Тема | Re: Suitable Index for my Table |
Дата | |
Msg-id | 1383590443.49877.YahooMailNeo@web162906.mail.bf1.yahoo.com обсуждение исходный текст |
Ответ на | Suitable Index for my Table ("Janek Sendrowski" <janek12@web.de>) |
Список | pgsql-general |
Janek Sendrowski <janek12@web.de> wrote: > I've got a table with many Values of the Type REAL. > These are my metric distances or my pivots to my sentences. > The table looks like this: > > ID INTEGER, distance1 REAL, distance2 REAL, distance3 REAL, > distance4 REAL, ..., distance24 REAL It should always raise a big red flag when you see column names with numeric suffixes like that. Usually these represent data which should be normalized out to another table, or possibly represented by a single column which is an array, hstore, or json type. > The range of the Value is in between 0 and 1. So it looks like > this 0.196 or 0.891 > > That my query > > WHERE value BETWEEN (distance1 - radius) AND (distance1 + radius) > WHERE value BETWEEN (distance2 - radius) AND (distance2 + radius) > WHERE value BETWEEN (distance3 - radius) AND (distance3 + radius) > WHERE value BETWEEN (distance4 - radius) AND (distance4 + radius) > ... Are those the WHERE clauses of four different queries, or did you mean for those to be four criteria on a single query? It would be better to show an actual, working query and table layout, so people have a more clear idea of the problem they are being asked to help solve. > Now I'm searching for a suitable index. A btree index on each distance column, maybe? -- Kevin Grittner EDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company
В списке pgsql-general по дате отправления: