Frequently Asked Questions

This FAQ is a living document. If you haven’t found the answer to your question here or anywhere else on our website, contact us via support@bupar.net, and we will get back to you as soon as we can. If you want to notify us of a bug in one of the bupaR-packages, please submit an issue on the respective GitHub-repository.

Create

What is an activity instance identifier, and when do I need one?

The activity instance identifier is a unique identifier for each activity instance, i.e. one specific execution of an activity for a specific case. If an activity exist of more than a single event, for example a start and a complete event, the activity instance identifier is needed to make sure these events are connected to each other.

When you don't have multiple events per activity instance, i.e. each activity only has one registrered timestamp, the activity instance identifier is redundant. In that case, you can use the activitylog object class, as specified in Scenario 1.

If your log has multiple events for the same activity instance, but there is no variable that can serve as an activity instance identifier to connect them, you can try assign_instance_id().

Creating the event log takes a lot of time. What can I do?

By default, eventlog() will perform some checks to make sure the provided mapping corresponds to the data model. This means that each value of the activity_instance_id should be connected to a single case_id and a single activity_id. For larger event logs, this does take some time. You can circumvent these by setting the argument validate = FALSE.

Alternatively, you might want to have a look at the different scenarios to decide whether you need an eventlog or activitylog object. The activitylog is a simplified object class that does not require special validation.

Visualize

How can I remove infrequent edges in a process map?

bupaR does not contain a magic slider that can simplify a process_map. We believe that process maps should always be a transparent visualization of the log, and simplifying should be done by the conscious use of filters. You might consider filter_infrequent_flows(), filter_trace_frequency() or filter_activity_frequency() for this job.

Can I use process maps in a Shiny app?

Yes, you can! Use the functions renderProcessMap() in the Server, and processMapOutput in the UI of your app. Make sure to use the width and height arguments to set proper dimensions for the map.

Can I save the process map to a file?

You can save a process map to a PDF, PNG, PS or EPS using export_map(). To do this, note that you should use the argument render = FALSE in the call to process_map(). Use the argument title to add a caption to the image, and width and height to adjust the file dimensions.