Hi,
I've got an interesting problem, which has stumped me for a week now.
I have a dataset that lists customer IDs and the product IDs for the products they currently own.
Simple example:
Tom, iphone
Tom, ipad
Tom, Minecraft Lego
Emma, iphone
Emma, ipad
Emma, guillotine
Emma, Ham sandwich
So the returned results would look like:
Iphone. Ipad. Etc
iphone
ipad
Minecraft lego
Guillotine
Ham sandwich
Where products coexist within a customer, just return a 1, but where they don't, return a zero.
I.e:
Ham sandwich and guillotine coexist within a customer, so their intersection would contain a 1
Iphone and ipad coexist within at least one customer, so their intersection would contain a 1
Ham sandwich and minecraft lego never coexist within the same customer, so their intersection would contain a zero.
I appreciate this is one of those things that may work better within SQL, but my client doesn't allow query analyser for their "normal" staff, (or even Access), so to be able to re-run this in the future, it's going to need to happen in excel.
Any ideas appreciated.
Tom.
(Edit) The actual data set is around 250k records once deduplicated across the combination of group and product
I've got an interesting problem, which has stumped me for a week now.
I have a dataset that lists customer IDs and the product IDs for the products they currently own.
Simple example:
Tom, iphone
Tom, ipad
Tom, Minecraft Lego
Emma, iphone
Emma, ipad
Emma, guillotine
Emma, Ham sandwich
So the returned results would look like:
Iphone. Ipad. Etc
iphone
ipad
Minecraft lego
Guillotine
Ham sandwich
Where products coexist within a customer, just return a 1, but where they don't, return a zero.
I.e:
Ham sandwich and guillotine coexist within a customer, so their intersection would contain a 1
Iphone and ipad coexist within at least one customer, so their intersection would contain a 1
Ham sandwich and minecraft lego never coexist within the same customer, so their intersection would contain a zero.
I appreciate this is one of those things that may work better within SQL, but my client doesn't allow query analyser for their "normal" staff, (or even Access), so to be able to re-run this in the future, it's going to need to happen in excel.
Any ideas appreciated.

Tom.
(Edit) The actual data set is around 250k records once deduplicated across the combination of group and product
Last edited: