Extract data from published dataset using external tools

[PA-JvdP]

Hi everyone,

I have a dataset which is published to the Power BI service in a premium workspace.

Using Python, I would like to extract all the data from this dataset to further analyze and work with it.

If I'm correct, this should be possible using the XMLA endpoint of the workspace. However, I'm not sure if its actually possible, or if the extracted data from the XMLA only concerns metadata about the workspace, datasets, relationships, etc.. I really need the most detailed level of the data, so not only high-level info. Is this possible?

Thanks in advance

[lbendlin · Accepted solution]

yes, it is possible. You can treat the dataset as if it were a SSAS tabular instance.

However, what you are trying to do is a mild red flag.  Instead of consuming the Power BI dataset you should probably be consuming its data sources directly.

[PA-JvdP]

Thank you! That's what I needed to know.

Why is this a red flag exactly? I understand it would be more efficient to consume the data sources themselves, but due to the large amount of processing steps this would take a long time. Other than efficiency, what are other concerns?

[lbendlin]

You are taking data out of a multidimensional cube and stuffing it into a flat table. That goes against the idea of what a cube is about.