Path Forecasting for a Process Instance
You can check the path forecasting for a process instance through the Monitor user interface in My webMethods. Path forecasting is based on aggregated historical data collected by Optimize and is available for currently running process instances that have been enabled for analysis.
When viewing the details for a process instance, you can select a forecast path and view the following estimated data for that path:
Estimated Completion Time - The estimated time of completion if the forecast path is taken.
Percentage Complete - The estimated percentage of completion for the process instance based on the selected forecast path’s Average Path Cycle Time.
Average Path Cycle Time - The average duration of the forecast path, calculated based on aggregated average step duration of the forecast path.
Average Process Cycle Time - The average execution duration of previously completed process instances. Process instances that were not fully completed do not contribute to the average cycle time.
Path Frequency - The frequency of the forecast path taken based on samples of historical data.
The estimated time of completion data is displayed for the entire process instance, not just for a single step. As the number of previously completed process instances increases, the accuracy of estimation also improves, because the estimation is based on a larger historical sample.
The path forecasting feature uses Optimize to provide estimations based on previously completed process instances. Optimize has a mode for calculating process and step statistical metrics, which is governed by the Optimize Analytic Engine’s Monitor Behavior Setting. This setting has three modes of operation:
All days are the same - all days of the week contribute to the same average.
Work days and weekend days -weekdays contribute to one average, while weekend days contribute to a separate average.
All days are different - each day of the week has its own average.
For more information on specifying statistical intervals, see Administering webMethods Optimize.
You should take into account the Optimize statistical mode of operation when checking estimated data for a forecast path. For example, if today is Tuesday and the Optimize statistical mode is “all days are different”, the estimation is based on past process instances completed on a Tuesday.
Note:
The Optimize Analytic Engine only calculates the process and step metric after the end of the day and averages do not include process instances for the current day.