Past patterns and established relationships can and do change invalidating the forecasts that inevitably are based on the extrapolation of such patterns and the continuation of relationships. The Monitoring module continuously checks for the difference between the actual and predicted values and provides a warning signal when such differences (i.e., the forecast errors) cease to be random. In such a case the user can re-forecast and/or take appropriate actions, if needed, to correct the situation.
PYTHIA supplies detailed reporting information on all aspects of data analysis and forecasting while also showing detail information about the accuracy of different types of predictions (original data without adjustments, adjusted data, statistical forecasts, judgmental overrides, budget objectives, and final forecasts). The reports can be customized and are exported in an Excel sheet for further usage.
PYTHIA was developed using Microsoft’s Visual Basic.NET 2008 while the Dundas Chart, for Basic. NET, was employed for the system implementation (the Dundas Chart was chosen for its advanced charting functionality and superior graphic options). Finally, the Microsoft SQL Server 2005 database is utilized by PYTHIA to store and retrieve the required information for the data analysis, forecasting and monitoring.