XML file integration in Azure Synapse Pipeline

Question

Hi there,

I'm trying to integrate a new XML file hosted in an Azure Data Lake Gen2 in Azure Synapse Workspace. I tried both way using a Copy Data activity and a Dataflow. Unfortunately this file is very complex with a big number of level in its hierarchy. When I run the data preview in the receptor of the data flow I encounter this message which blocks me for the end of the data flow :

This issue is unabling me the ability to flatten it afterward.

Any help would be appreciated,

Thanksfully

Answer 1

Hello @Quentin Chiffoleau ,
Thanks for the question and using MS Q&A platform.

As we understand the ask here is "how to mitigate a NaN error in Data Flow" .

NaN stands for Not a Number. This is different from the error which arises from toInteger('abcd') . NaN is more about floating point number types. This include Floats and Doubles.
NaN usually happens in computations involving division by 0, infinity, and imaginary numbers ( sqrt(-1 * N) ). There may also be cases where arithmetic overflow/underflow (value too big/small to store) could trigger a NaN.

When I run the data preview in the receptor of the data flow

Did you meant the Source Transformation, the start of the Data Flow? Did you mean the Sink Transforamtion, the end of the Data Flow?

Check out this similar case. Not all data stores support special values like infinity.
If this is on start side, try reading as string, not float.

Please do let me if you have any queries.

Thanks
Martin

---

• Please don't forget to click on or upvote button whenever the information provided helps you. Original posters help the community find answers faster by identifying the correct answer. Here is how
• Want a reminder to come back and check responses? Here is how to subscribe to a notification
  • If you are interested in joining the VM program and help shape the future of Q&A: Here is how you can be part of Q&A Volunteer Moderators

Answer 2

My bad, the error code is : Unable to parse NaN in preview data. Please try removing any number input that could produce NaN result.