Executing ML Python Scrip when Pressing Button

[imbusto]

Hello!

I have trained a linear regression model to predict the success of a promotion based of features like price, product or discount.

I want to develop a functionality in my PowerBI dashboard so when the user defines the values of these features, the success score is printed. Ideally, the success score would come from a Python script that normalizes user inputs, calls the trained model with those values, and returns the success score predicted by the model.

I am new to PowerBI and, although I spent some hours doing research, I still don't know if this functionality can be made. 

Please, let me know if there is any way to do this. If there is a way but does not involve using a python script, I would also be interested. 

Thanks in advance!

[lbendlin]

The Python visual in Power BI expects input data. Your Power BI semantic model can be influenced by report user choices (setting filters, clicking on visuals etc).  These influences (contexts) extend to the Python visual's intial data frame.

So - every time the users make filter choices your model could re-run.  If that is too often then there is also a deferral option ("apply filters now"). Note that your Python visual needs to render something after it finished the script.