Forum Discussion
Best Way to Import Accounting Statistics into a Visual?
I'm trying to import an accounting/finance report into Power BI, considering all the raw data is available (via an Oracle DB that I can access but not alter), but I'm having trouble figuring out which visual would be best to use. While the rest of the report uses table-centric information and was easy to put in visuals, there is one portion of the report that is just a rundown of statistics on two columns: Shipped and Ordered. Sections include average units per day, monthly target divided by number of business days in the month, difference between these numbers, and the percent variance, then doing something similar with monthly totals, YTD, Annualized, and YOY. Overall, it's 44 individual calculations over five sections put into a matrix format.
My first instinct is to make five cards and create a gigantic measure for each card that includes tabs and carriage returns with formulas embedded in it to approximate the matrix shape, but that seems like an exercise in futility. There must be an easier way,
and I'm just not thinking of it. I'd try to put together a table before importing the data, but I don't have access to building tables in Oracle, where the data lives. How would you visualize data like this?
1 Reply
- GebauerAnalyticFrequent Visitor
Another way I've found in my research is to create an index table of categories within Power BI, then write a gigantic SWITCH as the measure, distinguishing the criteria for each row. Again, seems like a messy solution, but it's better than my solution above. Are there actual programatic solutions, though?