XGBoost

This page lists the XGBoost (using the Scikit-Learn-like API) classes and functions that Bodo supports natively inside JIT functions.

Installing XGBoost

You will need to build XGBoost with MPI support from source. XGBoost version must be <= 1.5.1. Refer to XGBoost instructions about building requirements for more details. Then, build XGBoost with MPI support from source and install it in your Bodo environment as follows:

git clone --recursive https://github.com/dmlc/xgboost --branch v1.5.1
cd xgboost
mkdir build
cd build
cmake -DRABIT_BUILD_MPI=ON ..
make -j4
cd ../python-package
python setup.py install

xgboost.XGBClassifier

This class provides implementation of the scikit-learn API for XGBoost classification with distributed large-scale learning.

Methods

xgboost.XGBClassifier.fit

• xgboost.XGBClassifier.fit(X, y, sample_weight=None, base_margin=None, eval_set=None, eval_metric=None, early_stopping_rounds=None, verbose=True, xgb_model=None, sample_weight_eval_set=None, feature_weights=None, callbacks=None)
  
  Supported Arguments

  argument	datatypes
  X	NumPy Array or Pandas Dataframes
  y	NumPy Array or Pandas Dataframes

xgboost.XGBClassifier.predict

• xgboost.XGBClassifier.predict(X, output_margin=False, ntree_limit=None, validate_features=True, base_margin=None)
  
  

  Supported Arguments

  argument	datatypes
  X	NumPy Array or Pandas Dataframes

xgboost.XGBClassifier.predict_proba

• xgboost.XGBClassifier.predict_proba(X, ntree_limit=None, validate_features=True, base_margin=None)
  
  Supported Arguments

  argument	datatypes
  X	NumPy Array or Pandas Dataframes

Attributes

xgboost.XGBClassifier.feature_importances_

• xgboost.XGBClassifier.feature_importances_
  
  

Example Usage:

>>> import bodo
>>> import xgboost as xgb
>>> import numpy as np
>>> @bodo.jit
>>> def test_xgbc():
...   X = np.random.rand(5, 10)
...   y = np.random.randint(0, 2, 5)
...   clf = xgb.XGBClassifier(
...   booster="gbtree",
...   random_state=0,
...   tree_method="hist",
...   )
...   clf.fit(X, y)
...   print(clf.predict([[1, 2, 3, 4, 5, 6]]))
...   print(clf.feature_importances_)
...
>>> test_xgbc(X, y)
[1]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]

xgboost.XGBRegressor

This class provides implementation of the scikit-learn API for XGBoost regression with distributed large-scale learning.

Methods

xgboost.XGBRegressor.fit

• xgboost.XGBRegressor.fit(X, y, sample_weight=None, base_margin=None, eval_set=None, eval_metric=None, early_stopping_rounds=None, verbose=True, xgb_model=None, sample_weight_eval_set=None, feature_weights=None, callbacks=None)
  
  Supported Arguments

  argument	datatypes
  X	NumPy Array
  y	NumPy Array

xgboost.XGBRegressor.predict

• xgboost.XGBRegressor.predict(X, output_margin=False, ntree_limit=None, validate_features=True, base_margin=None)
  
  Supported Arguments

  argument	datatypes
  X	NumPy Array

Attributes

xgboost.XGBRegressor.feature_importances_

• xgboost.XGBRegressor.feature_importances_
  
  

Example Usage

>>> import bodo
>>> import xgboost as xgb
>>> import numpy as np
>>> np.random.seed(42)
>>> @bodo.jit
>>> def test_xgbc():
...   X = np.random.rand(5, 10)
...   y = np.random.rand(5)
...   clf = xgb.XGBRegressor()
...   clf.fit(X, y)
...   print(clf.predict([[1, 2, 3, 4, 5, 6]]))
...   print(clf.feature_importances_)
...
>>> test_xgbc(X, y)
[0.84368145]
[5.7460850e-01 1.2052832e-04 0.0000000e+00 4.2441860e-01 1.5441242e-04
 6.9795933e-04 0.0000000e+00 0.0000000e+00 0.0000000e+00 0.0000000e+00]