This is a simple copy of the table generation code from O'Neil et al.'s Star Schema Benchmark as described in the following paper:

Patrick O'Neil, Elizabeth (Betty) O'Neil and Xuedong Chen. "The Star Schema Benchmark," Online Publication of Database Generation program., January 2007. http://www.cs.umb.edu/~poneil/StarSchemaB.pdf

Changes with this copy :

    - Generating csv files instead of csv.
    - Changing the SEPARATOR to comme instead of |.

    make
    
    (customer.csv)
    dbgen -s 1 -T c
    
    (part.csv)
    dbgen -s 1 -T p
    
    (supplier.csv)
    dbgen -s 1 -T s
    
    (date.csv)
    dbgen -s 1 -T d
    
    (fact table lineorder.csv)
    dbgen -s 1 -T l
    
    (for all SSBM tables)
    dbgen -s 1 -T a


python dbgen.py
this will generate ssbRaw.csv:
	combining part.csv and customer.csv  with the lineorder.csv

These commands should generate the following files: customer.csv date.csv lineorder.csv part.csv supplier.csv

You can easily generate larger files by modifying the scale parameter (-s).

To generate the refresh (insert/delete) data set:

    dbgen -s 1 -r 5 -U 4

where "-r 5" specifies refreshin fact n/10000 "-U 4" specifies 4 segments for deletes and inserts. This will create the files create delete.[1-4] and lineorder.csv.u[1-4] with refreshing fact 0.05%.

This is strictly for research purposes.