One way is to use custom sql, to "arrange" your data properly, i.e. to have one record for every event (be it login or logout). Effectively this takes the first 3 columns from all records in your dataset, and then it appends at the bottom the "second" table comprising columns 1,4 & 5 (the structure of the whole table must be exactly the same). You could also add a field to distinguish between login/logout event, but anyway, having all restructured in this way allows you to plot all events over a time axis. Attached is a workbook.
Book2.twbx.zip 22.0 KB
In the back of my mind was that custom SQL might provide the solution, but way beyond my capability to create. With your example I'm sure I can build the general solution I'm after.
Thanks for taking the time to help me, and for the contribution to the Community.