What do you get when you combine the Separation Model + FM 14 placeholder text + ExecuteSQL + a million-record table + a variable array + Get(CalculationRepetitionNumber) + the Mod and Ceiling functions + a couple custom functions, with blazingly fast (local, LAN and WAN) performance thrown in for good measure?
This article w/ accompanying demo (fm-14-separated-data-mining.zip), that’s what.
Provide a data-mining interface to query a million-record table (cc_transactions) containing 20 years’ worth of credit card transactions. The client wants to be able to pick a date via a calendar widget, and see transaction info for that date, summarized by card type, transaction type and region.
This is a separated solution, with cc_transactions living in a file called z_data.fmp12, and a specific requirement for this project is to not make any schema changes to the data file. The data-mining will take place in a separate interface file… or in this case, two interface files, since we’re going to look at two methods. Continue reading