SparkSQL CSV的引用是模棱两可的

Question

我正在尝试读取SparkSQL 2.10中的一堆CSV文件，其自定义架构部分为双精度，部分为字符串，如下所示：

// Build the schema
val schemaStringS = "col1 col2"
val schemaStringD = "col3 col4 col5 col6"
val schemaStringS2 = "col7 col8"
val fieldsString = schemaStringS.split(" ")
  .map(fieldName => StructField(fieldName, StringType, nullable = true))
val fieldsString2 = schemaStringS2.split(" ")
  .map(fieldName => StructField(fieldName, StringType, nullable = true))
val fieldsDouble = schemaStringS.split(" ")
  .map(fieldName => StructField(fieldName, DoubleType, nullable = true))
val schema = StructType(fieldsString ++ fieldsDouble ++ fieldsString2)
// Read DataFrame
val input = sqlContext.read.schema(schema)
  .option("header", true)
  .csv("/files/*.csv")
  .toJavaRDD

这导致

Exception in thread "main" org.apache.spark.sql.AnalysisException: Reference 'col6' is ambiguous, could be: col6#0, col6#5.;
    at org.apache.spark.sql.catalyst.plans.logical.LogicalPlan.resolve(LogicalPlan.scala:264)
    at org.apache.spark.sql.catalyst.plans.logical.LogicalPlan.resolve(LogicalPlan.scala:158)
    at org.apache.spark.sql.catalyst.plans.logical.LogicalPlan$$anonfun$resolve$1.apply(LogicalPlan.scala:130)
    at org.apache.spark.sql.catalyst.plans.logical.LogicalPlan$$anonfun$resolve$1.apply(LogicalPlan.scala:129)
    at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
    at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:244)
    at scala.collection.Iterator$class.foreach(Iterator.scala:727)
    at scala.collection.AbstractIterator.foreach(Iterator.scala:1157)
    at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
    at org.apache.spark.sql.types.StructType.foreach(StructType.scala:96)
    at scala.collection.TraversableLike$class.map(TraversableLike.scala:244)
    at org.apache.spark.sql.types.StructType.map(StructType.scala:96)
    at org.apache.spark.sql.catalyst.plans.logical.LogicalPlan.resolve(LogicalPlan.scala:129)
    at org.apache.spark.sql.execution.datasources.FileSourceStrategy$.apply(FileSourceStrategy.scala:83)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$1.apply(QueryPlanner.scala:62)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$1.apply(QueryPlanner.scala:62)
    at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
    at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner.plan(QueryPlanner.scala:92)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2$$anonfun$apply$2.apply(QueryPlanner.scala:77)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2$$anonfun$apply$2.apply(QueryPlanner.scala:74)
    at scala.collection.TraversableOnce$$anonfun$foldLeft$1.apply(TraversableOnce.scala:144)
    at scala.collection.TraversableOnce$$anonfun$foldLeft$1.apply(TraversableOnce.scala:144)
    at scala.collection.Iterator$class.foreach(Iterator.scala:727)
    at scala.collection.AbstractIterator.foreach(Iterator.scala:1157)
    at scala.collection.TraversableOnce$class.foldLeft(TraversableOnce.scala:144)
    at scala.collection.AbstractIterator.foldLeft(Iterator.scala:1157)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2.apply(QueryPlanner.scala:74)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2.apply(QueryPlanner.scala:66)
    at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner.plan(QueryPlanner.scala:92)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2$$anonfun$apply$2.apply(QueryPlanner.scala:77)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2$$anonfun$apply$2.apply(QueryPlanner.scala:74)
    at scala.collection.TraversableOnce$$anonfun$foldLeft$1.apply(TraversableOnce.scala:144)
    at scala.collection.TraversableOnce$$anonfun$foldLeft$1.apply(TraversableOnce.scala:144)
    at scala.collection.Iterator$class.foreach(Iterator.scala:727)
    at scala.collection.AbstractIterator.foreach(Iterator.scala:1157)
    at scala.collection.TraversableOnce$class.foldLeft(TraversableOnce.scala:144)
    at scala.collection.AbstractIterator.foldLeft(Iterator.scala:1157)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2.apply(QueryPlanner.scala:74)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner$$anonfun$2.apply(QueryPlanner.scala:66)
    at scala.collection.Iterator$$anon$13.hasNext(Iterator.scala:371)
    at org.apache.spark.sql.catalyst.planning.QueryPlanner.plan(QueryPlanner.scala:92)
    at org.apache.spark.sql.execution.QueryExecution.sparkPlan$lzycompute(QueryExecution.scala:79)
    at org.apache.spark.sql.execution.QueryExecution.sparkPlan(QueryExecution.scala:75)
    at org.apache.spark.sql.execution.QueryExecution.executedPlan$lzycompute(QueryExecution.scala:84)
    at org.apache.spark.sql.execution.QueryExecution.executedPlan(QueryExecution.scala:84)
    at org.apache.spark.sql.execution.QueryExecution.toRdd$lzycompute(QueryExecution.scala:87)
    at org.apache.spark.sql.execution.QueryExecution.toRdd(QueryExecution.scala:87)
    at org.apache.spark.sql.Dataset.rdd$lzycompute(Dataset.scala:2547)
    at org.apache.spark.sql.Dataset.rdd(Dataset.scala:2544)
    at org.apache.spark.sql.Dataset.toJavaRDD(Dataset.scala:2557)
    at com.otterinasuit.spark.sensorlog.main.Main$.main(Main.scala:39)
    at com.otterinasuit.spark.sensorlog.main.Main.main(Main.scala)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:483)
    at com.intellij.rt.execution.application.AppMain.main(AppMain.java:147)

我尝试将文件与cat合并（仅适用于PoC）并避免使用CSV库（认为这可能是新Spark版本中的错误），但无济于事。

val input = sc.textFile("/csv/*.csv")
.map(line => line.split(",")).filter(row => !row.contains("col1")).map(x => Row(x))
val input2 = sqlContext.createDataFrame(input, schema)

我在常规数据帧和联接中遇到了这个问题，iirc 这可以通过指定列名、删除特定列或使用不同的联接来解决。但是，在这种情况下，我没有那个选择。

正如head -1 *.csv所证明的那样，所有文件中的所有标头都是相同的。我不明白为什么会发生这种情况。

Answer 1

fieldsString和fieldsDouble都指的是schemaStringS。

val fieldsString = schemaStringS.split(" ")
  .map(fieldName => StructField(fieldName, StringType, nullable = true))
//Changing from schemaStringS to schemaStringD
val fieldsDouble = schemaStringD.split(" ")
  .map(fieldName => StructField(fieldName, DoubleType, nullable = true))

所以，当你合并时

val schema = StructType(fieldsString ++ fieldsDouble ++ fieldsString2))

它抛出'col6' is ambiguous错误，