Demonstrating XGBoost Modin Interoperability

All the examples in this section are taken / adapted from https://xgboost.readthedocs.io/en/stable/python/python_intro.html

python

import xgboost as xgb
import modin.pandas as pd
import numpy as np

python

data_train = pd.DataFrame(np.arange(36).reshape((12,3)), columns=['a', 'b', 'c'])
label_train = pd.DataFrame(np.random.randint(2, size=12))
dtrain = xgb.DMatrix(data_train, label=label_train)

python

data_test = pd.DataFrame(np.arange(12).reshape((4,3)), columns=['a', 'b', 'c'])
label_test = pd.DataFrame(np.random.randint(2, size=4))
dtest = xgb.DMatrix(data_test, label=label_test)

python

param = {'max_depth': 2, 'eta': 1, 'objective': 'binary:logistic'}
param['nthread'] = 4
param['eval_metric'] = 'auc'

python

evallist = [(dtrain, 'train'), (dtest, 'eval')]
num_round = 10
bst = xgb.train(param, dtrain, num_round, evallist)

python

bst.save_model('0001.model')