Register now to learn Fabric in free live sessions led by the best Microsoft experts. From Apr 16 to May 9, in English and Spanish.
Hi,
I have a couple of dataflows which I link and combine in another dataflow. After saving the dataflow, I can succesfully connect to the dataflow in Power BI Desktop and build my report. For most data points this is not a problem. However, I have at least one data point, where the value of a calculated column changes from the dataflow web view to loading into Power BI Desktop.
As you can see in the screenshot, I want to check if a value in a column is empty or not. Here I create column "test" and the answer is clearly, that the column "sdmatstkonf.sdhza" is empty.
However, after connecting and loading in Power BI Desktop, column "test" says that "sdmatstkonf.hza" is NOT empty, even though "sdmatstkonf.hza" is indeed empty as can be seen in the column quality description.
I am sure that the data points are identical because I filtered the whole data set for this purpose to include only this data point.
This is a serious problem for me because I actually want to calculate the column next to the "test" column: "Echte hza". The calculation for this column follows the same logic. If sdmatstkonf.hza = "" then column x else column y.
Does anybody have an idea how this can happen? How can a custom column have different values after loading the dataflow?
Solved! Go to Solution.
Yep, it's inconsistent between cloud and desktop:
- https://community.powerbi.com/t5/Power-Query/Loading-dataflow-converts-null-to-empty/m-p/815361
I actually had the exact same case with a similar setup. Several dataflows (for Extract and Transform), where a certain column was TRUE, but in PBI Desktop it showed up as FALSE.
I was comparing empty and NULL strings, could it also be the case there's a difference in Power Query online vs. desktop?
It seems to be the case that there's a difference in Power Query online vs. desktop between empty and NULL (or somewhere in the background at least). I was able to "solve" my problem by extending the if statement:
Yep, it's inconsistent between cloud and desktop:
- https://community.powerbi.com/t5/Power-Query/Loading-dataflow-converts-null-to-empty/m-p/815361
Thanks @otravers, I did not find these posts when I searched for my problem. I did not realize it was due to the text format (which is also the format of my columns in question). I guess I can accept that as a solution as it explains the error and a workaround.
@v-easonf-msft I would hope Microsoft fixes the issue, but seeing that it is an old issue, I don't have much hope.
Hi, @Doni_No
That's really weird.
Will the same situation occur when creating a new ‘Test’ custom column in the integrated dataflow?
Best Regards,
Community Support Team _ Eason
Hi @v-easonf-msft,
thanks for the reply. As described in the answer to @nickyvv, I was able to solve my problem here, even though I did not solve the general behaviour of Power BI or really understand why it happened. I am also a bit unwilling to mark what I did as "solution", as it is more of a workaround I would say.
When I created a new test column in the integrated dataflow in Power BI Desktop with the same statement, the result was "sdmatstkonf.hza is empty". So I had two test columns with the exact same statement. The first executed in the dataflow was saying "sdmatstkonf.hza is NOT empty" and the other executed in Power BI Desktop said "sdmatstkonf is empty".
Covering the world! 9:00-10:30 AM Sydney, 4:00-5:30 PM CET (Paris/Berlin), 7:00-8:30 PM Mexico City
Check out the April 2024 Power BI update to learn about new features.
User | Count |
---|---|
58 | |
20 | |
19 | |
18 | |
9 |