我正在尝试使用Spark协同过滤实现推荐系统 .
首先,我准备模型并保存到磁盘:
MatrixFactorizationModel model = trainModel(inputDataRdd);
model.save(jsc.sc(), "/op/tc/model/");
当我使用单独的进程加载模型时,程序失败并出现以下异常:
Code:
static JavaSparkContext jsc ;
private static Options options;
static{
SparkConf conf = new SparkConf().setAppName("TC recommender application");
conf.set("spark.driver.allowMultipleContexts", "true");
jsc= new JavaSparkContext(conf);
}
MatrixFactorizationModel model = MatrixFactorizationModel.load(jsc.sc(),
"/op/tc/model/");
Exception:
线程“main”java.io.IOException中的异常:org.apache中的org.apache.hadoop.mapred.FileInputFormat.getSplits(FileInputFormat.java:324)中的文件:maprfs:/ op / tc / model / data不是.gark.rdd.HadoopRDD.getPartitions(HadoopRDD.scala:199)atg.apache.spark.rdd.RDD $$ anonfun $ partitions $ 2.apply(RDD.scala:239)at org.apache.spark.rdd.RDD $$ anonfun $在org.apache.spark.rdd.RDD.partitions(RDD.scala:237)的scala.Option.getOrElse(Option.scala:120)处分区$ 2.apply(RDD.scala:237) . aplet.spark.rdd.MapPartitionsRDD.getPartitions(MapPartitionsRDD.scala:35)atg.apache.spark.rdd.RDD $$ anonfun $ partitions $ 2.apply(RDD.scala:239)atg.apache.spark.rdd . RDD $$ anonfun $在org.apache.spark.rdd.RDD.partitions(RDD.scala:237)的scla.Option.getOrElse(Option.scala:120)处分区$ 2.apply(RDD.scala:237) .apache.spark.rdd.MapPartitionsRDD.getPartitions(MapPartitionsRDD.scala:35)atg.apache.spark.rdd.RDD $$ anonfun $ partitions $ 2.apply(RDD.scala:239)org.apache.spark.rdd .RDD $$ anonfun $分区$ 2 .apply(RDD.scala:237)位于org.apache.spark.rdd的org.apache.spark.rdd.RDD.partitions(RDD.scala:237)的scala.Option.getOrElse(Option.scala:120) . 位于org.apache.spark.rdd.RDD上的MapPartitionsRDD.getPartitions(MapPartitionsRDD.scala:35)$$ anonfun $ partitions $ 2.apply(RDD.scala:239)org.apache.spark.rdd.RDD $$ anonfun $ partitions $ 2.apply(RDD.scala:237)位于org.apache.spark.SparkContext的org.apache.spark.rdd.RDD.partitions(RDD.scala:237)的scala.Option.getOrElse(Option.scala:120) .runJob(SparkContext.scala:1952)atg.apache.spark.rdd.RDD $$ anonfun $ aggregate $ 1.apply(RDD.scala:1114)at org.apache.spark.rdd.RDDOperationScope $ .withScope(RDDOperationScope . scala:150)org.apache.spark.rdd.RDDOperationScope $ .withScope(RDDOperationScope.scala:111)位于org.apache.spark的org.apache.spark.rdd.RDD.withScope(RDD.scala:316) . rdd.RDD.aggregate(RDD.scala:1107)org.apache.spark.mll . 位于sun.reflect的sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)的com.aexp.cxp.recommendation.ProductRecommendationIndividual.main(ProductRecommendationIndividual.java:62)上的.mllib.recommendation.MatrixFactorizationModel.predict(MatrixFactorizationModel.scala:126) .nativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)位于org.apache.spark的java.lang.reflect.Method.invoke(Method.java:497)的sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) . deploy.SparkSubmit $ .org $ apache $ spark $ deploy $ org.apache.spark.deploy.SparkSubmit $ .doRunMain $ 1(SparkSubmit.scala:181)org.apache上的$ SparkSubmit $$ runMain(SparkSubmit.scala:742) . 在org.apache.spark.deploy.SparkSubmit.main(SparkSubmit . 斯卡拉)
我需要设置加载模型的配置吗?任何建议都会有很大的帮助 .