"polished" Northwind traders operational database
This is a PostgreSQL script for creating a traditional Northwind traders operational database sample.
"2x":
This dataset is actually made of 2 very similar Northwind datasets - this is why "2x". Northwind "orders" table is actually doubled in a continuous way - 2nd dataset's orders "continue" on the 1st datasets'. There is no overlapping in order dates. OrderItems stay the same - 1st dataset's 1st order items are the same as 2nd dataset's 1st order items, but they are also "doubled" up and each of the "twin tuples" references its own order.
"normalized":
2 Northwind datasets that I had were quite denormalized. I normalized most of the database "by the book". Tables have their primary keys, but also foreign keys defined. ER database model is included.
For even more polishing...
There are some inconsistencies due to e.g. versioning. E.g. there are cases where a customer obviously had their name changed and for that reason a new tuple was created. This is why customer name, address and shipping city where not removed from orders by normalizadion and were left as attributes. Feel free to find a way to resolve these "issues".
Probably even more minor inconsistencies can be encountered, but I decided this was already "clean" enough.
How to...
This is a script to c/p in your SQL editor and just click "run".
Doing this in a separate schema (instead of schema "public") is recomended. To do so, uncomment the first two lines of the script.