1. Components of process data
So far, we are already familiar with two important components in a process.
2. Cases and activities
The activities are the steps undertaken in the process, the actions.
And the cases are the objects flowing through the process on which the actions are performed.
In this lesson, we will learn about the distinction between activity instances and events, and the transactional lifecycle in a process. At the end of this chapter, you will be able to define your own event data objects in R!
3. Activity instances
When a specific action takes place for a specific case, this is called an activity instance: the execution of an activity in a specific context at a certain point in time. In this example, we see three activity instances, containing two different activities: Registration and X-ray, and two different patients: Emily and John.
In this example, we can see the time at which both the registration and the X-ray started. However, activities rarely happen instantly; they usually take some time. An activity instance therefore can consist of a series of events.
4. Events
If we take a closer look at the X-Ray activity execution for John, we would see that it was scheduled during registration at 9:51. It actually started at 10:42, and was finished at 10:58. Each of these time recordings is called an event. One or more events together can form an activity instance; three in this case.
The type of event -schedule, start, complete- is called the lifecycle status. In reality there can be many different statuses: assign, schedule, start, suspend, resume, abort. For simplicity, we will only use the most commons ones in this course - start and complete.
5. Event log
A log book of events is called an event log. For our current example, the event log looks as shown here. There are two registration instances, for which only a start event is recorded (the arrival of the patient), and there is 1 X-ray instance which consists of 3 events.
6. Resources
Another component of process data are the resources. As you can recall, resources are the actors in the process: they perform the activities. In this course, we always assume that a complete activity instance, that is all the related events, are executed by one and only one resource.
In our example, there are three resources: Mr. Owens and Mr. Fleming who performed the registration of John and Emily, respectively, and Dr. Russell who performed the X-ray.
7. Recap: event log
We now have all the ingredients for our process analysis: an list of events, with 6 different characteristics for each event: a case identifier, an activity label, an activity instance identifier, a timestamp, a lifecycle status, and a resource identifier.
8. Create event log object
In R, an event log object is created using the eventlog function from the bupaR package, which takes a data.frame with events as input, and returns a log object as output. In the data.frame, each of the ingredients we just saw should be present as a variable. In the eventlog function, we can specify each of the ingredients by referring to the variable names in the data.frame.
Once the eventlog object is created, we can start analyzing our process data in the next step!
9. Let's practice!
Now it's your turn. Let's see whether you can identify the different components in a process, and define event data accordingly!