dask / dask-examples

Easy-to-run example notebooks for Dask
https://examples.dask.org/
Creative Commons Attribution Share Alike 4.0 International
375 stars 228 forks source link

XGBoost example notebook uses deprecated dask-xgboost #223

Open avriiil opened 2 years ago

avriiil commented 2 years ago

This Dask Examples notebook uses the deprecated dask-xgboost rather than the native XGBoost integration.

Tweaking the notebook should be relatively straightforward by following the example code in this blog

import xgboost as xgb

# Create the XGBoost DMatrices
dtrain = xgb.dask.DaskDMatrix(client, X_train, y_train)
dtest = xgb.dask.DaskDMatrix(client, X_test, y_test)

# train the model
output = xgb.dask.train(
    client, params, dtrain, num_boost_round=4,
    evals=[(dtrain, 'train')]
)

# make predictions
y_pred = xgb.dask.predict(client, output, dtest)

From the dask-xgboost repo:

"Warning: Dask-XGBoost has been deprecated and is no longer maintained. The functionality of this project has been included directly in XGBoost. To use Dask and XGBoost together, please use xgboost.dask instead https://xgboost.readthedocs.io/en/latest/tutorials/dask.html."

avriiil commented 2 years ago

I can make these changes and submit a PR

jrbourbeau commented 2 years ago

Thanks for surfacing this @rrpelgrim. A PR that updates that example to use xgboost.dask instead of dask_xgboost would be very welcome. Let me know if you know have any questions about that process