We name the function non-missing instead of non-zero since zero is perfectly valid for XGBoost.

Create a dictionary of objects that can be pickled for function arguments.

Initialize data from a 2-D numpy matrix.

Generate the xgboost parameters which will be passed into xgboost library

Returns a dictionary of parameters and their default value of function fn. Only the parameters with a default value will be included.

Get the xgboost.XGBModel().fit() parameters

Get the parameters from xgboost.XGBModel().predict()

Get the special qid column from X if exists.

Must go before both pandas and cudf checks.

Decorate an objective function Converts an objective function using the typical sklearn metrics signature so that it is usable with ``xgboost

Parse an eval result string from the booster.

Get the XGBoost version from Python version file.

Convert array_like evaluation matrices into DMatrix. Perform sanity checks on the way.

Whether the numpy array has object dtype.

Wrapper to get an array.

Extract partitions from pyspark iterator. `append` is a user defined function for accepting new partition.

Basic test for the `ExtMemQuantileDMatrix`.

Check external memory with the quantile objective.

Test quantile loss with boosting random forest.

Copy files matching a glob pattern to a target directory.

Call shutil.copytree() with logging

Evaluate the Booster model with given dmatrixs. @param evalMatrixs dmatrixs for evaluation @param evalNames name for eval dmatrixs, used for check

Get the feature importances for gain or cover (average or total) @return featureImportanceMap key: feature index, values: feature importance score ba

Return the class name.

Get feature weights using the demo parser.

Extract the optional reference categories from the booster. Used for training continuation. The result should be passed to the :py:func:`pick_ref_

dummy definition for old mysys32.

Perform thread-safe prediction. @param data Flattened input matrix of features for prediction @param nrow The number of preditions to make

Synthetic dataset for learning to rank.

Get the number of non-missing values of DMatrix. @return The number of non-missing values @throws XGBoostError native error

Normalize UNIX path to a native path.

Compose the directory used for creating R package tar ball.

Parameters ---------- as_frame: A callable function to convert margin into DataFrame, useful for different df implementations.

assert that different workers count dosn't affect summ statistic's on root

Set feature names @param values feature names to be set @throws XGBoostError

Set feature types @param values feature types to be set @throws XGBoostError

Return SHAP values for an XGBoost model. .. warning:: This function is still working in progress. This function accepts either a :py:

Split feature index of node.

Return array interface json string for this Column

Get training result from parameters and data.

update one iteration @param iter iteration num @throws XGBoostError native error

create an empty histogram for i-th node