I want to create a trend measure that looks across a date period (Slicer). Counts the total number of rows in the dataset and then compares the first half of the data to the second half. That then would be the trend. In the attached PBX you will see the work I’ve done, i fell like I’m 70% of the way there, i can’t figure out the logic to get the middle of two dates?
The data represents the number of views each report gets over the month.
This is perfect… exactly what I was looking for.
I love how if there are 5 values it will count the first two and last two but leave the middle day out. I wasn’t sure how I was going to address uneven days in the filter. This actually does a good job of that.
With respect to the “weekends”. No need to address or add special logic, if the reports are viewed the days are counted. if the reports are NOT viewed the days are left out. I’m perfectly good with that.