Hi, Consider a retail stores credit card fee structure: Account type = A{1,...,a} - character Product type = B{1,..., b} - character Sub product type = C{1,..,c} - character Cost = D{1,...,d} - finite number of numeric costs Note that the length of B varies on the depending on the a_i in question and the length of C depends on the b_i (product type) in question. There is a specific cost associated with a given transaction on a given account type, product type and sub product type - that varies with transaction sizes. There are bands to this pricing and my objective is to highlight anomalies in this relationship. Bands example: 1) Transactions from $0 to $1000, we expect a fee of $10. 2) Transactions $1001 to $2000, we expect a fee of $15. Anomaly in 2) could be a transaction of $900 or $2100 with a fee of $15. These anomalies could be identified through building the aforementioned loop and highlighting the top and bottom 1% of transaction sizes for the given fee amount. I have struggled with building this in SAS. I have explored and tried tweaking numerous sets of code available on these forums without any luck. Please assist / advise - or is there any easier way to approach this problem in SAS?
... View more
