Traces from an Event Log in R

The common input for Process Mining tasks is the Event Log (and its common format the csv). Event Log files contain rows of events along with their timestamps (and of course their case Id).

However, you might need to have a file where each row is the sequence of the activities performed per case. Although, this is a quite common requirement, surprisingly, the current versions of ProM, do not provide such a feature! Maybe because it is so trivial…

Indeed, in R, after you import the csv file (let’s call it EventLog), you just need two lines of code!

> EventLog[with(EventLog,order(ID,Complete.Timestamp)),]

> tracesDF<- aggregate(EventLog$Activity,by=list(EventLog$ID),FUN=paste)