JPMML-SparkML-XGBoost

JPMML-SparkML plugin for converting XGBoost4J-Spark models to PMML.

Prerequisites

Apache Spark 2.3.X or 2.4.X.
XGBoost4J-Spark 0.7 or newer.

Installation

Enter the project root directory and build using Apache Maven:

mvn clean install

The build installs JPMML-SparkML-XGBoost library into local repository using coordinates org.jpmml:jpmml-sparkml-xgboost:1.0-SNAPSHOT.

Usage

The JPMML-SparkML-XGBoost library extends the JPMML-SparkML library with support for ml.dmlc.xgboost4j.scala.spark.XGBoostClassificationModel and ml.dmlc.xgboost4j.scala.spark.XGBoostRegressionModel prediction model classes.

Launch the Spark shell; use the --packages command-line option to include XGBoost4J-Spark, JPMML-SparkML and JPMML-XGBoost runtime dependencies, and the --jars command-line option to include the JPMML-SparkML-XGBoost runtime dependency:

spark-shell --packages ml.dmlc:xgboost4j-spark:0.90,org.jpmml:jpmml-sparkml:1.5.7,org.jpmml:jpmml-xgboost:1.3.15 --jars target/jpmml-sparkml-xgboost-1.0-SNAPSHOT.jar

Fitting and exporting an example pipeline model:

import ml.dmlc.xgboost4j.scala.spark.XGBoostClassifier
import org.apache.spark.ml.Pipeline
import org.apache.spark.ml.feature.RFormula
import org.jpmml.sparkml.PMMLBuilder

val df = spark.read.option("header", "true").option("inferSchema", "true").csv("Iris.csv")

val formula = new RFormula().setFormula("Species ~ .")
var classifier = new XGBoostClassifier(Map("objective" -> "multi:softmax", "num_class" -> 3))
classifier = classifier.set(classifier.numRound, 11)

val pipeline = new Pipeline().setStages(Array(formula, classifier))
val pipelineModel = pipeline.fit(df)

val pmmlBytes = new PMMLBuilder(df.schema, pipelineModel).buildByteArray()
println(new String(pmmlBytes, "UTF-8"))

License

JPMML-SparkML-XGBoost is licensed under the terms and conditions of the GNU Affero General Public License, Version 3.0.

If you would like to use JPMML-SparkML-XGBoost in a proprietary software project, then it is possible to enter into a licensing agreement which makes JPMML-SparkML-XGBoost available under the terms and conditions of the BSD 3-Clause License instead.

Additional information

JPMML-SparkML-XGBoost is developed and maintained by Openscoring Ltd, Estonia.