How to find coefficients of model in XGBoost Regressor?

Question

In XGBoost Regression to predict prices, How to get coefficients, intercepts of model? How to get summary of model like we get in Statsmodel for Linear regression? See below code

from xgboost import XGBRegressor

# fit model no training data
model = XGBRegressor()
model.fit(X_train, y_train)

# make predictions for test data
y_pred = model.predict(X_test)

print("R^2: {}".format(model.score(X_test, y_test)))
rmse = np.sqrt(mean_squared_error(y_test, y_pred))
print("Root Mean Squared Error: {}".format(rmse))

This is how I build the model and tried to get coefficients like this:

#print the intercept
print(model.intercept_)

AttributeError: Intercept (bias) is not defined for Booster type gbtree

print(model.coef_)

AttributeError: Coefficients are not defined for Booster type gbtree

Can someone please help me to solve this. Thanks.

score 8 · Answer 1 · edited Dec 13 '20 at 12:24

8

xgboost reference note on coef_ property:

Coefficients are only defined when the linear model is chosen as base learner (booster=gblinear). It is not defined for other base learner types, such as tree learners (booster=gbtree).

The default is booster=gbtree

edited Dec 13 '20 at 12:24

greybeard

2,249
8
30
66

answered Oct 31 '19 at 03:14

Vincent Snow

81
1
2

Welcome to SO! I already upvote your answer. It will be great if you add any reference to it. – David García Bodego Oct 31 '19 at 03:40
3

How do you access the coefficients when (booster=gblinear)? – Jan Janiszewski Jul 14 '20 at 10:00
@JanJaniszewski Simply pull out the coefficient array with model.coef_, they can be matched with the dataframe columns after doing this. – ninapberry Feb 19 '21 at 10:41

How to find coefficients of model in XGBoost Regressor?

1 Answers1