Advanced Set Analysis

[#1]

Please see this data table:

Brand	Item	Customer	Month
Eggo	Blueberry	Ryan	Jan
Eggo	Blueberry	Ryan	Feb
Waffle House	Blueberry	Ryan	Mar
Eggo	Strawberry	Ryan	Mar

I would like to write an analysis that gives me a count of Customers, By Month where the Customer switched the Brand but hanged on to the Item. For example in the Month of March Ryan switched to Waffle House but stayed with Blueberry item.

Thanks.

[Gysbert_Wassenaar]

And how do you handle two simultaneous changes?

Eggo/Blueberru->Waffle House/Blueberry + Eggo/Blueberry->Eggo/Strawberry

[#3]

We will not handle simultaneous changes. It will be always the first start within the time period selected. The base will always begin with Eggo. That will never change.

[#4]

I believe its easy to add the flag on the script level instead of set analysis if you data granularity is on Month level.

Set analysis won't work on the row by row comparison.