BTW – Foulke Management Corp has a 4-digit zip and so faked out the address block and gave a blank region. Obviously, in practice one could fiddle with some more calc fields to parse the state.

Calculations such as what you’ve suggested really do assist in finding commonality and outliers where combinations exist. Nice work! Techniques such as these not only work very well with this particular case, but are also readily portable to other challenges. Thanks for taking the time to share your thoughts.

Add …
Chrysler_Line_2: if(instr(“C”,Lines)>0,1,0)
Dodge_Line_2: if(instr(“D”,Lines)>0,2,0)
Jeep_Line_2: if(instr(“J”,Lines)>0,4,0)
Truck_Line_2: if(instr(“T”,Lines)>0,8,0)

These additional calculations allow is an interesting analysis extension effect represented by the value

Lines_Sum = Chrysler_Line_2 + Dodge_Line_2 + Jeep_Line_2 + Truck_Line_2.

This calculation yields a number which will be found in the range of 1 – 15. Each increment of Lines_Sum value represents a unique combination of Vehicle Lines closed *possible* in the data.

Thus … IF Lines_Sum = 6 then the lines closed can only Dodge and Jeep and no other combination. (The more anal-retentive readers (Does anal-retentive have a Hyphen??? (grin!)) will have noticed that the values assigned by the suggested additional calculated fields are powers of 2.)

Further, the absence of a value for Lines_Sum in the range of 1 – 15 in summary data means that the represented condition is NOT found in the data. For example, there are no 6 values (DJ only) of Sum_Lines and only two 9’s (CT only) found in the 789 closed dealerships. Also, multiple specific values for Lines_Sum could be selected to represent underlying combination condition. Dealerships having odd values of Lines_Sum must have the Chrysler line, etc.

I present this only for consideration as another tool for the analysis tool belt. Sometimes this technique can yield some interesting insights into data with which one is unfamiliar and which may contain hidden patterns that might have some significance.

I am looking forward to your next post.