K - type of key read from input topicM - type of message read from input topicU - type of model message written@FunctionalInterface public interface BatchLayerUpdate<K,M,U> extends Serializable
| Modifier and Type | Method and Description |
|---|---|
void |
runUpdate(org.apache.spark.api.java.JavaSparkContext sparkContext,
long timestamp,
org.apache.spark.api.java.JavaPairRDD<K,M> newData,
org.apache.spark.api.java.JavaPairRDD<K,M> pastData,
String modelDirString,
TopicProducer<String,U> modelUpdateTopic) |
void runUpdate(org.apache.spark.api.java.JavaSparkContext sparkContext,
long timestamp,
org.apache.spark.api.java.JavaPairRDD<K,M> newData,
org.apache.spark.api.java.JavaPairRDD<K,M> pastData,
String modelDirString,
TopicProducer<String,U> modelUpdateTopic)
throws IOException,
InterruptedException
sparkContext - Spark contexttimestamp - timestamp of current intervalnewData - data that has arrived in current intervalpastData - all previously-known data (may be null)modelDirString - String representation of path where models should be output, if desiredmodelUpdateTopic - topic to push models onto, if desired. Note that this may be null
if the application is configured to not produce updates to a topicIOException - if an error occurs during execution of the update functionInterruptedException - if the caller is interrupted waiting for parallel tasks
to completeCopyright © 2014–2018. All rights reserved.