@LDSamson I will give it a look. First impressions is that this PR actually reduces flexibility for us. We built the events tab off of the visits, so instead of using V0-Vx (or Visit 1 - Visit X), we were using C1D1-CXDY and there applicable names. This is also why we wanted the additional flexibility during merge_meta_with_data(). Further hard-coding the V0-Vx notation in the app wouldn't be a direction we would want to go. Our events tab looks like this:

I was not aware that you customized the events tab. This was previously only used for the compact timeline. On a side note, are you using the event_number for something specific? I don't think that was used in clinsight.

'Hard coding' was done for the small timeline but this can be changed. I think the rest is much more flexible in this way, also the normal 'Visit ...' notation is now completely customizable. You can create the same labels as you have now by filling in the table like this, if these patterns match the event_id in the raw data:

But you can still provide your own event_name (or other variables) if you prefer to do so, that did not change.

One advantage of this PR is that there can be many more visits, for example during follow-up. The number of visits do not not necessarily have to be hard coded in advance. And it provides a way to deal with unscheduled visits.

Marking as draft until custom visit labels can be used in the compact timeline again

Okay, I made improvements and restored the option to customize the compact timeline labels. @aclark02-arcus @jthompson-arcus can you have a look if it works for you?

The events tab in the metadata now looks like this:

Green columns are optional. From the orange ones, they can be used both but at least one needs to be available.
You can now just fill in the event_id column (previously event_label) and nothing else, and everything should be as expected.

Few more details;

• Only visits of which is_expected_visit is TRUE will show up in the timeline. Especially useful for 'unscheduled visits' which can happen at any time and should not show up here.
• Events will be created based on matching with event_id in the study data.
• The event_id will be taken as label for the compact timeline, but this can be overwritten by event_label_custom
• If event_id_pattern is used to match events, multiple events can match, and the labels will default to V0-V1-...-Vx for these events.
• The longer event_name will also default to event_id but can also be customized using event_name_custom. A sequence number will be added to the custom name if needed (add_sequence_to_name). If the column event_name_edc is available in the data, this value will be added as a suffix in brackets:
  

We need this construction mostly because we have studies with an indeterminate number of 'normal' visits and thus need some flexibility in creating event labels.