At 17:26 -0500 9/18/02, McIrvin wrote:
>OK,, I'll clarify:
>A query like:
>SELECT terms, count(*) AS count FROM searches GROUP BY terms HAVING
>count(*)>1 ORDER BY count DESC
>Is run on a table built by users searches on a site, so there are lots of
>duplicate entries, were trying to count the duplicates to determine the most
>popular search terms.
>The problem arises where you do a query like
>SELECT * FROM 'terms' WHERE term='test'
>And count the number of rows returned (with php or whatever) the issue is
>that the result from the count() function ends up being a little higher
>sometimes but by no more than 3 instances.
What happens if you replace count(*) by count(terms) throughout the
>----- Original Message -----
>From: "Paul DuBois" <paul@stripped>
>To: "McIrvin" <mcirvin@stripped>; <mysql@stripped>
>Sent: Wednesday, September 18, 2002 5:06 PM
>Subject: Re: count function issue
>> At 16:50 -0500 9/18/02, McIrvin wrote:
>> >Doing queries to determine the number of times search terms are used on a
>> >web site.
>> >Here is a sample sql query.
>> >SELECT terms, count(*) AS count FROM searches GROUP BY terms HAVING
>> >count(*)>1 ORDER BY count DESC
>> >But the results seem to be inaccurate, by no more that 3.
>> Sorry, it's difficult to parse that last sentence.
>> Also, since you don't indicate what you expect and what you actually
>> got, it's hard to say whether what you're seeing is to be expected.
> > >Anyone run into anything like this before?