java - 从嵌套的用户定义对象中创建 DataFrame

Question

我正在使用 MongoDB-Hadoop 连接器来读取具有嵌入文档的集合。

JSON 集合:PersonaMetaData

{
    "user_id" : NumberLong(2),
    "persona_created" : true,
    "persona_createdAt" : ISODate("2016-02-24T06:41:49.761Z"),
    "persona" : [{"persona_type" : 1,
            "created_using_algo" : "Name of the algo",
            "version_algo" : "1.0",
            "createdAt" : ISODate("2016-02-24T06:41:49.761Z"),
             "persona_items": {"key1":"value1", "key2": "value2"} }]
}

我创建了以下类来表示集合中的数据

 class Persona_Items implements Serializable
 {
   private int key1;
   private String key2;
   // Getter/Setter and constructor 
 }

 class Persona implements Serializable
 {
   String persona_type;
   String created_using_algo
   String version_algo
   long createdAt;
   List<Persona_Items> listPersonaItems; 
     // Getter/setter and constructor
 }

 class PersonaMetaData implements Serializable
 {
   long user_id;
   boolean persona_created;
   long persona_createdAt;
   List<Persona> listPersona;

   // Getter/setter and constructor 
 }

然后将其用作

 // RDD representing the complete collection
 JavaPairRDD<Object, BSONObject> bsonRdd = sc.newAPIHadoopRDD(inputConfig,
                     com.mongodb.hadoop.MongoInputFormat.class,
                     Object.class, BSONObject.class);

 // Get RDD of PersonaMetaData
 JavaRDD<PersonaMetaData> metaDataSchemaJavaRDD =
 bsonRdd.map(new Function<Tuple2<Object, BSONObject>, PersonaMetaData >() {
                         @Override
                         public PersonaMetaData  call(Tuple2<Object, BSONObject> objectBSONObjectTuple2)
                                 throws Exception { // Parse the BSON object and return a new PersonaMetaData object }

 // Convert into DataFrame

 dataFrame= sqlContext.createDataFrame(metaDataSchemaJavaRDD,
                     PersonaMetaData.class);

异常

scala.MatchError: io.abc.spark.schema.PersonaMetaData @31ff5060 (of class io.abc.spark.schema.PersonaMetaData ) at org.apache.spark.sql.catalyst.CatalystTypeConverters$StructConverter.toCatalystImpl(CatalystTypeConverters.scala:255) at org.apache.spark.sql.catalyst.CatalystTypeConverters$StructConverter.toCatalystImpl(CatalystTypeConverters.scala:250) at org.apache.spark.sql.catalyst.CatalystTypeConverters$CatalystTypeConverter.toCatalyst(CatalystTypeConverters.scala:102) at org.apache.spark.sql.catalyst.CatalystTypeConverters$ArrayConverter.toCatalystImpl(CatalystTypeConverters.scala:169) at org.apache.spark.sql.catalyst.CatalystTypeConverters$ArrayConverter.toCatalystImpl(CatalystTypeConverters.scala:153) at org.apache.spark.sql.catalyst.CatalystTypeConverters$CatalystTypeConverter.toCatalyst(CatalystTypeConverters.scala:102) at org.apache.spark.sql.catalyst.CatalystTypeConverters$$anonfun$createToCatalystConverter$2.apply(CatalystTypeConverters.scala:401) at org.apache.spark.sql.SQLContext$$anonfun$9$$anonfun$apply$1$$anonfun$apply$2.apply(SQLContext.scala:500) at org.apache.spark.sql.SQLContext$$anonfun$9$$anonfun$apply$1$$anonfun$apply$2.apply(SQLContext.scala:500) at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244) at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244) at scala.collection.IndexedSeqOptimized$class.foreach(IndexedSeqOptimized.scala:33) at scala.collection.mutable.ArrayOps$ofRef.foreach(ArrayOps.scala:108) at scala.collection.TraversableLike$class.map(TraversableLike.scala:244) at scala.collection.mutable.ArrayOps$ofRef.map(ArrayOps.scala:108) at org.apache.spark.sql.SQLContext$$anonfun$9$$anonfun$apply$1.apply(SQLContext.scala:500) at org.apache.spark.sql.SQLContext$$anonfun$9$$anonfun$apply$1.apply(SQLContext.scala:498)

Not having any list in the class runs without any issues.

Answer 1

正如 Inferring the Schema Using Reflection 中明确指出的那样Spark SQL, DataFrames and Datasets Guide 部分

Spark SQL does not support JavaBeans that contain nested or contain complex types such as Lists or Arrays.