Message-ID: <460A8865.3070005@fmed.uba.ar>
Date: Wed, 28 Mar 2007 12:23:17 -0300
From: Gerardo Herzig <gherzig@fmed.uba.ar>
User-Agent: Mozilla Thunderbird 1.0.7 (X11/20050923)
MIME-Version: 1.0
To: pgsql-sql@postgresql.org
CC: andreas.kretschmer@schollglas.com
Subject: Re: union with count?
References: <460A75ED.6070501@fmed.uba.ar>
	<20070328142444.GA20168@a-kretschmer.de>
In-Reply-To: <20070328142444.GA20168@a-kretschmer.de>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Thanks! But now i have another problem related with count():

select page_id, word, word_position, count(page_id) from (select * from 
search_word('word1', 'table1') union search_word('word2', 'table2')) foo 
group by page_id;

and gives me "foo.word must appear in GROUP clause or be used in an 
aggregate function"

And i want to group by page_id only, because that is what i need to 
count. Tips here?

Thanks again man.
Gerardo

>am  Wed, dem 28.03.2007, um 11:04:29 -0300 mailte Gerardo Herzig folgendes:
>  
>
>>Hi dudes, i have the following question
>>i have 2 tables with the same format, and i want to know if is there a 
>>possibility of using some sort of count(*) for achieving this result:
>>select a from table1 where (condition) union select a from table2 where 
>>(condition), count(a) group by a
>>
>>The idea is having how many times (condition) is true for both tables.
>>
>>Im not shure im explaining myself clearly, my english is not the best 
>>(as you can see ;)
>>    
>>
>
>You can use a subselect, a simple example:
>
>test=# select *, count(1) from (select 1 union select 2 union select 3) foo group by 1;
> ?column? | count
>----------+-------
>        1 |     1
>        2 |     1
>        3 |     1
>(3 rows)
>
>
>Andreas
>  
>