evaluate XGBoostModel with a RDD-wrapped dataset
evaluate XGBoostModel with a RDD-wrapped dataset
NOTE: you have to specify value of either eval or iter; when you specify both, this method adopts the default eval metric of model
the dataset used for evaluation
the name of evaluation
the customized evaluation function, null by default to use the default metric of model
the current iteration, -1 to be null to use customized evaluation functions
the average metric over all partitions
whether to output raw margin
Predict result with the given test set (represented as RDD)
Predict result with the given test set (represented as RDD)
test set represented as RDD
whether to use external cache for the test set
Predict result with the given test set (represented as RDD)
Predict result with the given test set (represented as RDD)
test set represented as RDD
the specified value to represent the missing value
Predict leaf instances with the given test set (represented as RDD)
Predict leaf instances with the given test set (represented as RDD)
test set represented as RDD
the name of the column storing the raw prediction value, either probabilities (as default) or raw margin value
Save the model as to HDFS-compatible file system.
Save the model as to HDFS-compatible file system.
The model path as in Hadoop path.
Thresholds in multi-class classification
produces the prediction results and append as an additional column in the original dataset NOTE: the prediction results is kept as the original format of xgboost
produces the prediction results and append as an additional column in the original dataset NOTE: the prediction results is kept as the original format of xgboost
the original dataframe with an additional column containing prediction results
append leaf index of each row as an additional column in the original dataset
append leaf index of each row as an additional column in the original dataset
the original dataframe with an additional column containing prediction results
class of the XGBoost model used for classification task