Type Mismatch Error

[ChrisHedges]

Hey

I've got an error with very little help from the error report:

Last refresh failed: 13/10/2023, 10:45:54 am
There was an error when processing the data in the dataset.
Data source error: The following system error occurred: Type mismatch. Table: Stock Item Availability.
Cluster URI: WABI-SOUTH-EAST-ASIA-redirect.analysis.windows.net
Activity ID: 8e449420-2241-4533-9a9a-8a345da7a1a2
Request ID: 5e536643-b98f-9ff1-f53f-61721fe52656
Time: 2023-10-13 00:15:54Z

I am able to refresh and publish my report fine, however when I try and refresh on PowerBI online I am given the below error.  THe table I get an error with is called Stock Item Availability. On desktop it refreshed fine (which is coming via a dataflow, also refreshing fine). Although when I go to refresh online I get a Type Mismatch error in the Stock Item Availability table, however I am not given any further information about columns etc.

[SaiTejaTalasila]

Hi,

Turn on column quality on the dataflow/dataset and check for any errors. From my experience, most of the time, the error occurs because of a null value in the date/time column. So, you have to replace it with some date.( you can use dates from before 1900 in place of null. Later on, you can identify them and you can filter.)

[mp9531943781]

@ChrisHedges  Did you get a proper solution for this error ?

I'm facing the same issue with my semantic model.

[HaNPT]

I'm having the same error, it only shows the table name, and I don't make any change with the data type, still keep the same type as the source, does anyone figure out the reason?

[collinq]

Hey @HaNPT ,

You will have to go to the file itself, to the table mentioned and look at the data in the columns.  This is not very intuitive I know but it almost suredly is something in one or more fields in one of the columns that is not quite right.  Often, this is a date field but not always.  If your data is quite tall (lots of rows) you might just go to the visual page, create a slicer visual and stick in each thing field by field and look for what is unique.

[oscarrudnas]

The issue might be related to having a datetime / timestamp with timezones:

"2023-08-10 14:07:18.000 UTC" is possible on Power BI Desktop, but it seems to convert that to datetime "2023-08-10 14:07:18" in the desktop app, and it apparently can't do that in Power BI Service.

So removing the timezone from the source or maybe possible with Power Query might help (haven't tried that, since I could change in the source).

Hope that helps someone.

[collinq]

Hi @ChrisHedges ,

The algorithms on the Service are much less forgiving than in the Desktop.  Type Mismatch is usually about having a field with a certain type where some of the data in the column are not that type.  
The fix is to go back to desktop and look and see if you have any "ANY" columns.  Make those a specific data type.  Also, check that you have DateFields that have only valid dates in them - that seems to be a consistent issue for folks.

You can also look and see if any columns have an "error" in the value field (this is hard to do sometimes if you have many rows and many columns).

You can also confirm that you have no fields that have "list" or "record" in the value.

[joku50]

Hi @collinq ! I had the same issue and your response here helped me fix the issue that I struggled for days to resolve 🙂 Thank you so much and I hope you have a wonderful day!