There are some common issues/problems that many Power BI users will encounter at some point or another during their course of using Power BI. In this blog article, I attempt to catalog the most common, reoccurring issues that are repeatedly posted to the forums. Before you post, make sure that it isn’t already a known and solved common problem by consulting the list below.
And, looking for input and improvements. Did I miss any common issues/problems? Did I miss any great blog articles that can resolve any of these problems (likely!)
The totals in your table/matrix are wrong
This is almost certainly caused by what is referred to as “the measure totals problem”. This one is extremely common. See this post that explains it, Dealing with Measure Totals
You are trying to get year-to-date, previous year-to-date, month-to-date or do something using any of the “time intelligence” DAX functions.
First, if you are trying to use the time intelligence functions in DAX and do not have a separate calendar table, you need to create one using CALENDAR or CALENDARAUTO. Then you would do things like this:
Also, if you are still failing with DAX time intelligence functions or have a specific circumstance that they do not cover or just am tired of the whole magical “black box” that are time intelligence functions that do not have explicit control over, then just do it with filters because that’s all they are, Time Intelligence the Hard Way
You want to perform a calculation between two rows of a table, especially if you are looking for a calculation between the “previous” or “next” row
You are likely going to want to use the EARLIER function or create a variable using VAR that stores the current row’s value and then essentially “lookup” the previous or next row using LOOKUPVALUE or MAXX(FILTER(…),…) or MINX(FILTER(…),…). See this article on Mean Time Between Failure (MTBF) which uses EARLIER.
You are having an issue with RANKX
There are tricks to using RANKX. See these blog articles:
You are having trouble with a Many-To-Many relationship
Try to get rid of your many-to-many relationship by inserting a bridge table of DISTINCT values between the two tables and using the column from your bridge table in your matrix. Another approach is to create a combination key using concatenation of 2 or more columns.
You need the aggregate of a measure
Another common scenario is that you wish to get the average or some other aggregation of a measure for a group of “things”. So think of the circumstance where you are using a measure to calculate something but now you want to know the average or sum or max or min of that measure over some grouping of rows. This is the measure aggregation problem. See the blog article about that here, Design Pattern Groups and Supergroups