Tobit with yl and yu varied by observations

[kota7]

Thanks for this repo!
I wonder if it is possible to extend the BoostingRegressor to work with the data where censoring points yl, yu are varied by observations.

[fabsig]

Thank you for using KTBoost!

Yes, this is possible. You basically have to add the option to use arrays instead of numbers for yl and yu in the TobitLossFunction. I currently don't have time to work on this. Contributions are welcome.