If you wind up using the query, and it 'makes a profit' (makes things easier, or
simpler to maintain) for your employer, then the original author wouldn't be out
of line in asking for some compensation, would he?
PMG wrote:
>
> First of all, thanks for your reply. There is a real world need for the type
> of query I am looking for - I think it's called a Pareto query. In our case,
> we have a database that tracks production problems. Each type of problem has
> a problem_code. So, management is interested in seeing which problem_codes
> are most frequent, and they, being management, are not interested in the
> minutiae, but only in the ones that account for 80% of all production
> problems. According to the Pareto 80/20 principle, approx 20% of the
> problem_codes will account for approximately 80% of the problems.
>
> Presently, I am implementing this the old-fashioned way -- i.e. non-SQL
> procedural -- by inserting group data into a table, then going back to the
> first row, and calculating the percentage of the total for the first group,
> then the next, etc. Finally I sort.
> It works, but I was hoping that there is a nice little query (that an
> ordinary guy like me can understand) that will do the same thing.
>
> Regarding your query, it will take me a month of Sundays to try to figure
> out what is happening! Re your intimation for $100, is this a rule-based
> hint or a cost-based hint?
>
Received on Sun Dec 13 1998 - 10:09:38 CST
