Forum Discussion
How to remove error rows ? Remove error option not working
Hi,
How to remove error rows ? Remove error option not work refer below SS
Row number 19
when i click remove error its not working
Error Still showing in data table
please help how to remove this error row
Hi lavdeep,
This one stumped me for a while I was. I would transform data to eliminate errors and it just didn't take. Eventually this is how I solved it:
1) Make sure you are in the Power Query Editor
2) Go to the "Transform" tab
3) Select the column you want to modify
4) Look for "Replace Values", and more importantly, the little down arrow next to it. It's in the sub-section titled "Any Column"
5) Click the down arrow next to "Replace Values" and you'll see another option, "Replace Errors"
If you click that it should produce a dialogue which allows you to replace any errors in that column (or columns, if you select multiple) with a value of your choosing.
11 Replies
Hi lavdeep,
This one stumped me for a while I was. I would transform data to eliminate errors and it just didn't take. Eventually this is how I solved it:
1) Make sure you are in the Power Query Editor
2) Go to the "Transform" tab
3) Select the column you want to modify
4) Look for "Replace Values", and more importantly, the little down arrow next to it. It's in the sub-section titled "Any Column"
5) Click the down arrow next to "Replace Values" and you'll see another option, "Replace Errors"
If you click that it should produce a dialogue which allows you to replace any errors in that column (or columns, if you select multiple) with a value of your choosing.This solution didn't work for me UNTIL I changed my order of Steps.
I had a Filtered Rows step and if I tried 'Removing Errors' or 'Replacing Errors' afterwards, then it failed.
But when I started applying this step and then Filtering, it worked!Would love to know why, but in the meantime I'll just enjoy it and stop banging my head against the table 😅
Thank you SOOOO much!
If the errors are in several columns, then you can select all the affected columns and then use then select "Remove Errors"
Hi,
the provided solution is not working for me. Any other suggestions?
JacekI appreciate this solution, however, unfortunately it is still not solving my problem. I have a row deep in the dataset (row 3239, far beyond the 1000 row preview) that gives an error across all columns. The error shows the following:
Expression.Error: We cannot convert the value "University of Arizon..." to type Table.
Details:
Value=University of Arizona
Type=[Type]I have tried removing rows with errors, replacing rows with errors, and even deleting individual columns until theres nothing left. I believe this is truly a defective row and want to get rid of it entirely but nothing is working. Any other suggestions? Could I write a conditional to remove any rows that contain "University of Arizona"?
I had a similar problem with an excel dataset that I was bringing into the Power Query editor. After countless tries, I went into the source data and applied an IFERROR formula to result in "null" and then reloaded the file - which worked perfectly.
Just in case this helps someone else, I was able to fix this by changing the ApiVersion from 14 to 15 in the Source line
Another possible resolution (and, I think, the best one if you're able to do it) is to identify where in your query the errors are being generated and fix that.
In my case, I was invoking a custom function that generated multiple table columns. I ended up going through each one of those tables and found the table that had the error row (every field in that row returned Error, and no Remove Errors or Replace Errors step would resolve it). I then went back to the query from which the custom function was created and populated the values for the problematic table. Sure enough, there were a substantial number of entries that had Error in one of the columns. When I used Replace Error in that query, it fixed the issue in the query that was calling the Invoke Function step.
It was somewhat time consuming, and it may not always be possible to do this, but if you can, it's definitely the best approach, because it's actually addressing a problem with the data, not with PBI.
