I think it can be done in a few steps. It takes some time however.
1. Make a query A with customer number and "A" for all distinct customernumbers having bought product A.
2. Same up to L.
3. Make a query with your customer table as primary entry and an outer join to query A through L. Output is customernumber, followed by 12 times a letter or a NULL. Say 6,,,D,,,,,,, and 26,A,B,C,,,,,,,,,
4. Make a query to convert the 12 values into one string products_bought, replacing NULL with space. So 6," D " and 26,"ABC ".
5. A group-by-query of this result by products_bought.
It isn't difficult, it's step by step, so you have ample opportunity to check the intermediate results.
Kees
I am working with one table of data. The data has customer numbers (ex 1,2,3,4) and also products bought by those customer numbers (ex a,b,c,d). The customer number may have bought one or more products. Each purchase is a different entry or row of information.
I have around 100,000 rows.
I want to find some information in the simplest way possible.
1. How many and what customers have each of the possible combinations.
For example (customers 6,1000,and 16000 have only product d while customers 26, 88, and 4678 all have products a,b,c.)
There are 12 products of which a customer may have purchased one or more.
Any ideas of a simple way to get to this information and a sample query?
Thanks

Chowhound
Comic Vine
GameFAQs
GameSpot
Giant Bomb
TechRepublic