Sergio Calderón Pérez-Lozao
Sergio Calderón Pérez-Lozao
Thank you very much @bolliger32 your help was incredible and now I can show nbdiff-web between base branches and remotes but I have still a problem with nbmerge-web :( This...
Now I'm using this as alternative solution: ```python import pandas as pd import numpy as np from sklego.datasets import load_heroes from category_encoders import LeaveOneOutEncoder df_heroes = load_heroes(as_frame=True).replace([np.inf, -np.inf], np.nan).dropna() #...
Do you need any help with this issue? Is it something missed on my part?
Hi! My Spark version is 2.4, I´ll also try your notebook code.
OK thank you!!
Thanks for your fast response @dsgibbons! I think it is (look for example [this line](https://github.com/shap/shap/blob/master/shap/plots/_partial_dependence.py#L89)).
The problem is when I run the model in cluster mode (not local). It´s a Yarn Cluster as I mentioned before.
Sorry, I had to tell to you that I also tested with seed parameter and it's not reproducible :(
@TomAugspurger can you reply please? :(
Sorry @TomAugspurger ;( Hi @DigitalPig, The point is that if I build two xgboost models with exactly the same parameters it doesn´t return the same model because the importances are...