effective_xgboost_book
effective_xgboost_book copied to clipboard
mattharrison
Hello, I noticed a typo on Page 59. In the third line from the bottom, "The XBGoost library provides" should be corrected to "XGBoost". Thank you for your attention to...
```py from sklearn import model_selection params = {'reg_lambda': [0], 'learning_rate': [.1, .3], 'subsample': [.7, 1], 'max_depth': [2, 3], 'random_state': [42], 'n_jobs': [-1], 'n_estimators': [200]} xgb2 = xgb.XGBClassifier(early_stopping_rounds=5) cv = (model_selection.GridSearchCV(xgb2,...
This use of round as a variable name shadows the built-in round() function within the scope of the for loop. It's generally considered poor practice to use the names of...
The `violinplot` function in seaborn doesn't need a `kind` parameter because it's already clear what kind of plot you want—a violin plot. You may have confused this with the `catplot`...
Hi 🐍 Matt Harrison, I'm thoroughly enjoying your book on XGBoost, but I noticed what might be data leakage during hyperparameter tuning. Specifically, on page 46, 47, 48 and 49,...
Hello Matt Harrison, I've been enjoying your book so far. However, I noticed some inconsistencies in the decision tree visualizations, specifically on page 53. While this seems to occur in...
Page 27 of the book - had to change the code to this: fig, ax = plt.subplots(figsize=(8, 4)) tree.plot_tree(stump, feature_names=['value'], filled=True, class_names=list(stump.classes_), ax=ax) with class_names param needing a list
In order to avoid the standard issues w/ different versions and compatability; is there something which documents which software and which versions are used?
xg_helpers library mentioned/referenced in book but I can't locate it in repo nor book.
It seems that using the seed 42 with xgboost 1.7.6 with the sample data used in section 8.3 returns a score of `0.7535911602209945` which is different from the value on...
