DB2 SQL 根据权重选择不同的数据

Question

I have a table of categories, each of which has a weighting applied to it.

I have a table of Products and Categories that they appear in, and want to select the Category with the highest weighting for each Product.

I have tried to use Distinct, but it doesn't seem to want to work with sort.

My basic query would be:

select a.product, a.category from prodcattable a
join categorytable b on a.category=b.category
order by product, weighting desc

Where a product appears in multiple categories, I just want to return one row, with the product and the category that ranks the highest.

This needs to work in DB2 as I need to run it on an iSeries.

I am hoping that I can use the result to compare against a master product table to see if the category has changed, without having to read through the whole master product table, and for each product find the category with the highest weighting. With over 75,000 products, that's a lot of queries to execute.

Answer 1

One method uses row_number() :

select pc.*
from (select p.product, p.category,
             row_number() over (partition by p.product order by weighting desc) as seqnum
      from prodcattable p join
           categorytable c
           on p.category= c.category
     ) pc
where seqnum = 1;