You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have the train_data with X = ['C', 'C', 'C', 'C', 'C', 'A', 'C', 'B', 'C', 'B', 'C', 'B', 'A', 'C', 'A', 'A', 'B', 'A', 'A', 'B'] and y=[0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1]. I put the data into the model and get
(Prior set 0.55 here because in CatBoostEncoder, prior set as the mean value of y)
After that, I used CatBoostEncoder in sklearn and convert the cat_feature into the numerical
I got catf=[8.25, 4.125, 2.75, 5.8125, 4.65, 8.25, 3.875, 8.25, 3.32142857, 11.625, 2.90625, 12.75, 11.625, 4.25, 12.75, 13.3125, 13.3125, 10.65 , 8.875, 10.65]. (multiply 15 here because the scale value in model1 is 15)
And then feed the data into model but get difference result, does there exist any extra process when deal with the category during training?
The text was updated successfully, but these errors were encountered:
andrey-khropov
changed the title
Get difference tree result when convet cat_features to numerical value
Get difference tree result when converting cat_features to numerical values
Apr 10, 2024
catboost version: catboost 1.2
Operating System: macos 14.4.1
CPU: M1
GPU: no
I have the train_data with X = ['C', 'C', 'C', 'C', 'C', 'A', 'C', 'B', 'C', 'B', 'C', 'B', 'A', 'C', 'A', 'A', 'B', 'A', 'A', 'B'] and y=[0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1]. I put the data into the model and get
(Prior set 0.55 here because in CatBoostEncoder, prior set as the mean value of y)
After that, I used CatBoostEncoder in sklearn and convert the cat_feature into the numerical
I got catf=[8.25, 4.125, 2.75, 5.8125, 4.65, 8.25, 3.875, 8.25, 3.32142857, 11.625, 2.90625, 12.75, 11.625, 4.25, 12.75, 13.3125, 13.3125, 10.65 , 8.875, 10.65]. (multiply 15 here because the scale value in model1 is 15)
And then feed the data into model but get difference result, does there exist any extra process when deal with the category during training?
The text was updated successfully, but these errors were encountered: