I have a calculated member that looks like this:
[Measures].[Transactions]
Once processed, my powerbi has 6 rows, which is correct:
Then I went back to the calculated member and added the 1- to the formula:
This is what my powerbi looks like now:
It's still showing Emp Group Miami but it's showing every single DimEmp, even though there isn't a relationship between the two.
This same behavior can be seen in the Visual Studio cube browser.
Regarding the sctucture I have two dim tables: DimEmpGroup and DimEmp. In PowerBi, I applied filter (DimEmpGroup equals Miami). Powerbi automatically displays DimEmp Miami_Joe, Miami_Mary, and Miami_Bill since they belong to DimEmpGroup Miami.
As soon as I add this subtraction to the calculated member, The relationship between DimEmpGroup and DimEmp breaks. Powerbi still has filter DimEmpGroup Miami applied (as before), but now I'm seeing every DimEmp. The only way I can see the correct rows is by also applying filter to DimEmp.
I assume that this is not how one subtracts in MDX, but I haven't found a solution.
Question: how can I subtract a fixed number?
Related
I am learning Power BI Desktop. I have a project already loaded for which I have created a
pie chart.
As you can see, the data is currently divided into two years: 2022 and 2021.
I also have data for previous years that I want to include in the chart. Unfortunately, none of these years has a high number. So I want to group them all into an "Other" category.
I've already tried using the RANKX function as described here (where I copied and pasted the formula from the comment and replaced the field name with my own) and here (where I went through the article step by step using my own field names). I used the exact format shown in both the comment and the article:
Rank = RANKX('Table','Table'[Percent],,DESC,Dense)
(Table is the table name I am using and Percent is the column I am trying to rank.)
When I added the new column to a table, though, PowerBI automatically gave me a sum calculation. This calculation gave me a 1 in each column. When I asked it not to give me a sum, it said it could not display the visual and took me back to the calculation screen.
EDIT: Thanks for your help Kevin! This is the table I tried to draw from. I am an idiot sometimes.
EDIT 2: Now I am trying to use a different field in the formula:
Rank = RANKX('FactClaimActivity','FactClaimActivity'[DirectIncurredLoss_ITD])
(This is the original field that I tried to base Percent on)
My table looks like this. (Yes, I know that 2016, 2017, 2019 and 2020 have negative values and 2014 and 2015 have $0 values. I just want to show what the data is.)
EDIT 3: This time I am also getting an error when I type in the formula, which says "A single value for column 'DirectIncurredLoss_ITD' in table 'FactClaimActivity' cannot be determined. This can happen when a measure formula refers to a column that contains many values without specifying an aggregation such as min, max, count, or sum to get a single result."
I do not understand why it wants a single result for a formula that ranks the data points based on this field. That seems like it defeats the whole purpose of the formula.
Yet when I use the formula, I get the same result as last time! This column is in the Fixed decimal number format, and it does have many different values.
EDIT: I've found the answer
New Category = IF('FactClaimActivity'[Rank]<=2,'FactClaimActivity'[LossYear],"Other")
The formula will help you get the correct result.
I have to divide two measures from two different tables. I have created a measure in Table A & created measure-2 in Table B.
When I use matrix visual in Power BI by taking date field in columns and region in rows (for table A&B), I can see the both table values are correct as I am expected.
Ex: Table A 2017-Q1 value by measure1 is 29.2, Table B 2017-Q1 value by measure1 is 2.9.
I have to divide both measures and I need to show the value (divide%) in TableA along with Measure1.
Unfortunately I tried in multiple ways by forming relationship b/w two tables also, But not getting the expected result i.e., 29.2/2.9 we should get 10% but instead of that getting 3%.
Without knowing your data model, it's hard to give a reasonable answer.
https://learn.microsoft.com/en-us/dax/related-function-dax
Your best change of understanding what happens is to learn up on relations, and changes them when needed. The documentation is a great starting point.
Unrelated data plotted in a visual of different data will always aggregate since there is no relation to split your values. The value of 3% is correct, your assumption that you want 10% as an outcome is not valid for your situation.
If you link the dates of table A and the dates of table B to a seperate Calendar, it all would work.
To preface this, I'm fairly experienced in Excel and VBA but new to PowerBI and more than a bit confused.
I have a flat table with a [creationdate]-, [Prio] (Priority (1,2,3)) and a calculated [Days Open] column, among many irrelevant others. I need to create a chart that displays the average days a case was open by priority of the case.
To display the average "days required" per (opening-) month for the past 18 months, I created the following measure:
Prio 1 = CALCULATE(AVERAGE('SourceName'[Days Open]),'SourceName'[Prio]=1)
Then I used that as a value, and used the [creationdate] as the x-axis. (Later I changed the x-axis to a new date table linked to [creationdate] without it making a difference.) To display this as monthly averages, I used the hierarchy limited to years and months, and went down one level in the chart.
Something seemed off so I checked first in Excel, then in the data source in PowerBI and yep: The averages in the PowerBI chart are complete bullshit.
Where did I go wrong? I assume it has something to do with the date hierarchy... So I created a date table as recommended (which....why?!) and linked it. That didn't make a difference.
Meanwhile in the data panel if I filter by the date column and calculate the average with the filtered selection of numbers externally, everything works as expected, so its not like there's a date formatting issue.
Do I have to create a calculated column with something akin to
DATE(YEAR([DateColumn]),MONTH([DateColumn]),1)
, then use that as the x-axis without the hierarchy, and hope nobody cares about the day in the label? Or is there something wrong with the measure used? I'm completely lost.
Two-line graph with slicers: Everything looks fine until date selections are made.
I figure that this failure has something to do with the line graph and date functionality. Date selections work together in a very weird way. Selecting only 2019 months and then only 2020 months, both lines disappear.
If the date slicers are going to work together, I should expect inclusive behavior, rather than "NOT 2020 AND NOT 2019"!
See relationships.
How do I get the behavior intended on one two-line graph?
Thanks!
I agree with #teylyn, sometimes less is more.
The problem is how Power Bi works and how your measures are calculated. You have two averagex-measures, both pointing to same date table. In Power BI, two separate filters from a single dimension can't do "union", they are laid down on top of each other. You are telling that first I want to filter by 2019 November and after the filter has been evaluated, you tell that you want to filter 2020 January. Power BI adds new filter to already filtered dataset, but the dataset doesn't contain anything that match both filters, so the end result is empty.
One way to get around this problem is to create separate date-tables for filtering purposes for blue and green. Here is an example of the data model needed:
Then you update the date-filters in second tab. Set blue date filter to one date-dimension and green date filter to another. This way they aren't filtering the same table and the graph should work the same as before.
I'm new to the DAX world and I just faced a conceptual misunderstanding.
While I was trying to debug a measure, I found a unexpected (for me) behaviour with the ALL function. I wrote a simple measure COUNTROWS(ALL(Table)) but to my surprise, when I sliced the data by a column, the same number did not appear in each cell in the values field, ignoring the filter context in the way I was expecting:
So, what am I missing?
Since there are no BUYs on e.g. 01/08/2017, it's not even evaluating the measure. Since that cell doesn't correspond to any rows in the original table, it's not doing any computations.
If you had a BUY and a SELL on the same day, it should show up in both columns on that day.