Quantile Regression using Gradient Boosting Tree Model GBT

[caltang]

Has anyone had experience with running a quantile regression using gradient boosting tree model GBT in Alteryx before? 

Or is Python Tool the only way to get it up? 

[acarter881 · Accepted solution]

Hey, @caltang.

It sounds like something that can be done only in the Python tool or the R tool. I don't think editing the Boosted Model macro would be worth it, but I suppose it depends on your familiarity with R.

For Python, I believe https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html is a good example.

I don't believe there's a way to do this natively in Alteryx without custom code in Python or R. You may be better off trying it in an IDE, such as Visual Studio Code, then translating it to Alteryx if it needs to be there.

[caltang · Accepted solution]

Yeah I’ve got something up and running on Python already. Trying to port it to Alteryx and add interface tools over it so that end users don’t have to code. 

i guess that’s my only option.

[acarter881]

Hello, @caltang.

I see what you mean. I've created GUIs before with tkinter: https://docs.python.org/3/library/tkinter.html and Qt Designer: https://doc.qt.io/qt-6/qtdesigner-manual.html.

They're fun if you want to then save those GUIs as executable files so people can run an application just by double clicking an icon on their desktop.

If you do create one in Designer with the Python tool, it may be cool to share some screenshots or your experience creating it.

[caltang]

Very cool! I'll study that more. Thanks for recommending! 

Also, sure! Building on Python in Alteryx then packaging it as a Standard Macro kinda makes it into its own tool. I'll share it once I finish!