From: | Justin <justin(at)emproshunts(dot)com> |
---|---|
To: | dmp <danap(at)ttc-cmc(dot)net> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: how do you write aggregate function |
Date: | 2008-03-09 18:43:28 |
Message-ID: | 47D42FD0.9030801@emproshunts.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
If i use two arrays how do i append new entries into these arrays.
array_append only allows for 1 dimension.
what other methods are there to add new entries to the array.
dmp wrote:
> Array appends are usually a performance hit, as you said. I'm not sure
> though with
> PostgreSQL. Why not try it with two arrays and see what happens. At
> least you would
> reducing the single array and the eliminating the append.
>
> danap.
>
>> I got the aggregate function for weighted average done. I finely
>> left alone for more than 10 minutes to actual get it written. It
>> takes 2 value input Weight and the Value. it will sum the weighted
>> entries to create the constant then does the normal formula, but does
>> not percentage number but averaged number. A simple change on the
>> return line it can do percentages.
>>
>> I did a couple of things a little odd . instead of doing a multi
>> dimensional array i did a one dimensional array where the 1st row is
>> Weight and the 2nd row is Value. This made the loop through the
>> array look stupid.
>> I tested it across 50,000 records with a group by it took 3.3 seconds
>> to run.
>>
>> without the group by clause performance is terrible taking several
>> minutes just to do the sfunc part. 371,563ms
>>
>> The Array seems to have performance hit any advice? It could be
>> the way i'm appending to the Array which has a performance hit as the
>> array gets bigger and bigger ?
>
>
From | Date | Subject | |
---|---|---|---|
Next Message | Mitchell D. Russell | 2008-03-09 18:49:11 | Re: v8.3 + UTF8 errors when restoring DB |
Previous Message | dmp | 2008-03-09 18:17:46 | Re: how do you write aggregate function |