Wrong Totals in Power BI

Yes, you heard it right Power BI still hasn't figured out how to show the correct totals. One of the most common questions in the Power BI community is related to it. So, let's figure out what can you do to correct your totals.

I am using Sample Superstore data to start getting rid of the implicit measures that have already been created. I am creating a matrix with categories and sub-categories. Let's include the total sales, total profit, and the total number of orders received (if you are wondering how these measures are created then it's just the sum of sales, profit, and distinct count of order ID respectively).

Protip- One of the best practices involves creating a separate table for measures (stand-alone table). In this case, we only have 5 measures but what if you have 70-80 measures all in the fact table.

If you critically note the total for sales and quantity sold shows the correct total but if you sum up the total number of orders for furniture it doesn't sum up to 1764. Also, if we look at the total at the end for several orders it is 5009 which exactly matches the distinct count of the order ID.

What's actually going on?? Let's debug this so the total on the matrix is the total where we haven't defined the level on which the calculation needs to be done so by default it shows me the distinct count of order id. There are multiple ways to correct this the most common approach will be using SUMX and values.

Let's see how it works. I am creating another measure using SUMX and Values. We will define the level as a sub-category and the measure will look something like this.


Now let's get back to our matrix. You can see now the total is corrected. But is it the correct number?


With the new measure, you are creating an iterated sum on the level of the subgroup. 1764 in the furniture category shows the distinct count of order IDs and there can be two products in a category with the same order id. Thus, 1764 only shows the distinct order ID associated with a sub-category which is correct while on the other hand, 1984 can include the multiple order ID (can show wrong results). 

If you are using the above method be cautious about the level you want to sum. Another method is not so common but you can create calculated columns(follow the link to learn more about calculated columns). This is something a BI developer doesn't prefer because calculated columns can come up with their own disadvantages.


Thanks for Reading Let's connect on LinkedIn. For more such blogs and pro tips do follow us




Comments

Post a Comment

Popular posts from this blog

7 Commonly used DAX expressions in Power BI

Image
In today's blog, we will focus on making the Power BI journey more riveting in nature. We will target the Data Analysis Expressions (DAX) in Power BI. Let's get started by getting a gist of it. DAX is a sort of language in Power BI which helps to create measures, calculated columns, and customized tables by using various functions and operators. It makes the Power BI more interactive in nature which can result in better visualization. We will focus on the seven most commonly used DAX expressions in Power BI. These can vary according to your needs but according to my novice experience, these expressions prove out to be very common and nifty in nature. So let's jump to the expressions. We are using sample superstore data for every expression. Sum and SumX - As the name suggests sum is responsible to add all the values in the column. But if you spot the difference between the two the latter one is responsible for returning the sum of expressions for each row in the table. For ...
Read more

SQL Joins

Image
Are you aware of the fundamentals of a Set theory which entails a critical element for our database i.e. the concept of joins? Why do we need Joins in the database? Joins are the fundamental element if you want to combine various tables using a common or related column between them. Joins are mainly used in every analytical and business intelligence tool.  Let's take a deep dive into the fundamentals of joins which is apparent by using a Venn diagram. There are mainly four types of joins- Inner, Left, Right, and Full outer join. Let's suppose we have two datasets A and B. When we focus on inner join it basically means the focus will be on the intersection or common columns present in both datasets. Similarly, the left join focuses on the dataset that is available in A and the common columns between both data sets. Right join is quite identical to left join but the main difference is it takes the data from B along with the common columns present between both data sets. Lastly, f...
Read more

Copying Bookmarks from one Power BI report to another

Image
Let's think of a scenario, where you want to copy the bookmarks from one report to another. Most obvious approach is to just do a copy paste of the bookmarks. What's wrong with this approach? This approach only works for all visuals but not for bookmarks and field parameters.  If you are not familiar with basics of bookmarks and field parameters do refer to the beginners guide for bookmarks  and introduction to field parameters . Then how do you copy the bookmarks? Power BI enhanced report format (PBIR) for Power BI Project files (PBIP) will help you in achieving this. Let's check it out, I have 2 reports one contains the bookmark called Bookmarks PBIR Test (origin) and other one is Rolling Average PBIR Test (destination) .  Before we get started, you have to enable Power BI Project save option under preview features. Once enabled, restart Power BI desktop. There is a TMDL icon appearing on the left pane. What is TMDL and what's in it for me? There's a lot of possi...
Read more