我有一个文本文件,我在火花数据框中将其作为CSV文件读取。 现在加入后我写一个函数时,为了选择列,我得到异常。
这是我的代码加载csv文件
val df = sqlContext.read.format("csv").option("header", "true").option("delimiter", "|").option("inferSchema","true").load("s3://trfsdisu/SPARK/FinancialLineItem/MAIN")
val df1With_ = df.toDF(df.columns.map(_.replace(".", "_")): _*)
val column_to_keep = df1With_.columns.filter(v => (!v.contains("^") && !v.contains("!") && !v.contains("_c"))).toSeq
val df1result = df1With_.select(column_to_keep.head, column_to_keep.tail: _*)
val df2 = sqlContext.read.format("csv").option("header", "true").option("delimiter", "|").option("inferSchema","true").load("s3://trfsdisu/SPARK/FinancialLineItem/INCR")
val df2With_ = df2.toDF(df2.columns.map(_.replace(".", "_")): _*)
val df2column_to_keep = df2With_.columns.filter(v => (!v.contains("^") && !v.contains("!") && !v.contains("_c"))).toSeq
val df2result = df2With_.select(df2column_to_keep.head, df2column_to_keep.tail: _*)
import org.apache.spark.sql.expressions._
val windowSpec = Window.partitionBy("LineItem_organizationId", "LineItem_lineItemId").orderBy($"TimeStamp".cast(LongType).desc)
val latestForEachKey = df2result.withColumn("rank", rank().over(windowSpec)).filter($"rank" === 1).drop("rank", "TimeStamp")
这是我的架构
latestForEachKey.printSchema()
root
|-- DataPartiotion: string (nullable = true)
|-- LineItem_organizationId: long (nullable = true)
|-- LineItem_lineItemId: integer (nullable = true)
|-- StatementTypeCode_1: string (nullable = true)
|-- LineItemName_1: string (nullable = true)
|-- LocalLanguageLabel_1: string (nullable = true)
|-- FinancialConceptLocal_1: string (nullable = true)
|-- FinancialConceptGlobal_1: string (nullable = true)
|-- IsDimensional_1: boolean (nullable = true)
|-- InstrumentId_1: string (nullable = true)
|-- LineItemSequence_1: string (nullable = true)
|-- PhysicalMeasureId_1: string (nullable = true)
|-- FinancialConceptCodeGlobalSecondary_1: string (nullable = true)
|-- IsRangeAllowed_1: string (nullable = true)
|-- IsSegmentedByOrigin_1: string (nullable = true)
|-- SegmentGroupDescription_1: string (nullable = true)
|-- SegmentChildDescription_1: string (nullable = true)
|-- SegmentChildLocalLanguageLabel_1: string (nullable = true)
|-- LocalLanguageLabel_languageId_1: string (nullable = true)
|-- LineItemName_languageId_1: integer (nullable = true)
|-- SegmentChildDescription_languageId_1: string (nullable = true)
|-- SegmentChildLocalLanguageLabel_languageId_1: string (nullable = true)
|-- SegmentGroupDescription_languageId_1: string (nullable = true)
|-- SegmentMultipleFundbDescription_1: string (nullable = true)
|-- SegmentMultipleFundbDescription_languageId_1: string (nullable = true)
|-- IsCredit_1: string (nullable = true)
|-- FinancialConceptLocalId_1: string (nullable = true)
|-- FinancialConceptGlobalId_1: string (nullable = true)
|-- FinancialConceptCodeGlobalSecondaryId_1: string (nullable = true)
|-- FFAction_1: string (nullable = true)
df1result.printSchema()
root
|-- LineItem_organizationId: long (nullable = true)
|-- LineItem_lineItemId: integer (nullable = true)
|-- StatementTypeCode: string (nullable = true)
|-- LineItemName: string (nullable = true)
|-- LocalLanguageLabel: string (nullable = true)
|-- FinancialConceptLocal: string (nullable = true)
|-- FinancialConceptGlobal: string (nullable = true)
|-- IsDimensional: boolean (nullable = true)
|-- InstrumentId: string (nullable = true)
|-- LineItemSequence: string (nullable = true)
|-- PhysicalMeasureId: string (nullable = true)
|-- FinancialConceptCodeGlobalSecondary: string (nullable = true)
|-- IsRangeAllowed: boolean (nullable = true)
|-- IsSegmentedByOrigin: boolean (nullable = true)
|-- SegmentGroupDescription: string (nullable = true)
|-- SegmentChildDescription: string (nullable = true)
|-- SegmentChildLocalLanguageLabel: string (nullable = true)
|-- LocalLanguageLabel_languageId: integer (nullable = true)
|-- LineItemName_languageId: integer (nullable = true)
|-- SegmentChildDescription_languageId: integer (nullable = true)
|-- SegmentChildLocalLanguageLabel_languageId: integer (nullable = true)
|-- SegmentGroupDescription_languageId: integer (nullable = true)
|-- SegmentMultipleFundbDescription: string (nullable = true)
|-- SegmentMultipleFundbDescription_languageId: integer (nullable = true)
|-- IsCredit: boolean (nullable = true)
|-- FinancialConceptLocalId: integer (nullable = true)
|-- FinancialConceptGlobalId: integer (nullable = true)
|-- FinancialConceptCodeGlobalSecondaryId: string (nullable = true)
|-- FFAction: string (nullable = true)
这是我收到错误的地方
val dfMainOutput = df1result.join(latestForEachKey, Seq("LineItem_organizationId", "LineItem_lineItemId"), "outer")
.select($"LineItem_organizationId", $"LineItem_lineItemId",
when($"StatementTypeCode_1".isNotNull, $"StatementTypeCode_1").otherwise($"StatementTypeCode").as("StatementTypeCode"),
when($"LocalLanguageLabel_1".isNotNull, $"LocalLanguageLabel_1").otherwise($"LocalLanguageLabel").as("LocalLanguageLabel"),
when($"FinancialConceptLocal_1".isNotNull, $"FinancialConceptLocal_1").otherwise($"FinancialConceptLocal").as("FinancialConceptLocal"),
when($"FinancialConceptGlobal_1".isNotNull, $"FinancialConceptGlobal_1").otherwise($"FinancialConceptGlobal").as("FinancialConceptGlobal"),
when($"IsDimensional_1".isNotNull, $"IsDimensional_1").otherwise($"IsDimensional").as("IsDimensional"),
when($"InstrumentId_1".isNotNull, $"InstrumentId_1").otherwise($"InstrumentId").as("InstrumentId"),
when($"LineItemLineItemName_1".isNotNull, $"LineItemLineItemName_1").otherwise($"LineItemLineItemName").as("LineItemLineItemName"),
when($"PhysicalMeasureId_1".isNotNull, $"PhysicalMeasureId_1").otherwise($"PhysicalMeasureId").as("PhysicalMeasureId"),
when($"FinancialConceptCodeGlobalSecondary_1".isNotNull, $"FinancialConceptCodeGlobalSecondary_1").otherwise($"FinancialConceptCodeGlobalSecondary").as("FinancialConceptCodeGlobalSecondary"),
when($"IsRangeAllowed_1".isNotNull, $"IsRangeAllowed_1").otherwise($"IsRangeAllowed").as("IsRangeAllowed"),
when($"IsSegmentedByOrigin_1".isNotNull, $"IsSegmentedByOrigin_1").otherwise($"IsSegmentedByOrigin").as("IsSegmentedByOrigin"),
when($"SegmentGroupDescription_1".isNotNull, $"SegmentGroupDescription_1").otherwise($"SegmentGroupDescription").as("SegmentGroupDescription"),
when($"SegmentChildDescription_1".isNotNull, $"SegmentChildDescription_1").otherwise($"SegmentChildDescription").as("SegmentChildDescription"),
when($"SegmentChildLocalLanguageLabel_1".isNotNull, $"SegmentChildLocalLanguageLabel_1").otherwise($"SegmentChildLocalLanguageLabel").as("SegmentChildLocalLanguageLabel"),
when($"LocalLanguageLabel_languageId_1".isNotNull, $"LocalLanguageLabel_languageId_1").otherwise($"LocalLanguageLabel_languageId").as("LocalLanguageLabel_languageId"),
when($"LineItemName_languageId_1".isNotNull, $"LineItemName_languageId_1").otherwise($"LineItemName_languageId").as("LineItemName_languageId"),
when($"SegmentChildDescription_languageId_1".isNotNull, $"SegmentChildDescription_languageId_1").otherwise($"SegmentChildDescription_languageId").as("SegmentChildDescription_languageId"),
when($"SegmentChildLocalLanguageLabel_languageId_1".isNotNull, $"SegmentChildLocalLanguageLabel_languageId_1").otherwise($"SegmentChildLocalLanguageLabel_languageId").as("SegmentChildLocalLanguageLabel_languageId"),
when($"SegmentGroupDescription_languageId_1".isNotNull, $"SegmentGroupDescription_languageId_1").otherwise($"SegmentGroupDescription_languageId").as("SegmentGroupDescription_languageId"),
when($"SegmentMultipleFundbDescription_1".isNotNull, $"SegmentMultipleFundbDescription_1").otherwise($"SegmentMultipleFundbDescription").as("SegmentMultipleFundbDescription"),
when($"SegmentMultipleFundbDescription_languageId_1".isNotNull, $"SegmentMultipleFundbDescription_languageId_1").otherwise($"SegmentMultipleFundbDescription_languageId").as("SegmentMultipleFundbDescription_languageId"),
when($"IsCredit_1".isNotNull, $"IsCredit_1").otherwise($"IsCredit").as("IsCredit"),
when($"FinancialConceptLocalId_1".isNotNull, $"FinancialConceptLocalId_1").otherwise($"FinancialConceptLocalId").as("FinancialConceptLocalId"),
when($"FinancialConceptGlobalId_1".isNotNull, $"FinancialConceptGlobalId_1").otherwise($"FinancialConceptGlobalId").as("FinancialConceptGlobalId"),
when($"FinancialConceptCodeGlobalSecondaryId_1".isNotNull, $"FinancialConceptCodeGlobalSecondaryId_1").otherwise($"FinancialConceptCodeGlobalSecondaryId").as("FinancialConceptCodeGlobalSecondaryId"),
when($"FFAction_1".isNotNull, $"FFAction_1").otherwise($"FFAction").as("FFAction"))
.filter(!$"FFAction".contains("D"))
dfMainOutput.write
.format("csv")
.option("quote", "\uFEFF")
.option("codec", "gzip")
.save("s3://trfsdisu/SPARK/FinancialLineItem/output")
下面的是我的例外
org.apache.spark.sql.AnalysisException: cannot resolve 'CASE WHEN (`IsRangeAllowed_1` IS NOT NULL) THEN `IsRangeAllowed_1` ELSE `IsRangeAllowed` END' due to data type mismatch: THEN and ELSE expressions should all be same type or coercible to a common type;
我在加载CSV时没有提到任何类型。
我将两者都加载为没有架构的csv文件但是在一个IsRangeAllowed_1中是字符串,而在另一个中它是BooleanType
有了这个,还有一个我想问的问题。
我们如何在数据框输出中删除默认分隔符并将自定义分隔符与分区和gzip压缩一起放置?
dfMainOutput.rdd.saveAsTextFile("s3://trfsdisu/SPARK/FinancialLineItem/output")
答案 0 :(得分:1)
对于第一个问题,我,e WHEN THEN和ELSE表达式应该是相同类型或普通类型;
但此处IsRangeAllowed
为Boolean
,IsRangeAllowed_1
为String
。因此,将其中一列转换为String或Boolean。所以代码更改可能是
import org.apache.spark.sql.types.DataTypes
when($"IsRangeAllowed_1".isNotNull, $"IsRangeAllowed_1")
.otherwise($"IsRangeAllowed".cast(DataTypes.StringType))
.as("IsRangeAllowed")
我们如何在数据框输出中删除默认分隔符并将自定义分隔符与分区和gzip压缩一起放置?
DataFrame
可以直接与delimiter
和codec
一起保存,而无需调用基础rdd
,例如dfMainOutput.rdd
。即:
dfMainOutput.write
.format("csv")
.option("delimiter", "!")
.option("codec", "gzip")
.save("s3://trfsdisu/SPARK/FinancialLineItem/output")
编辑:根据concat_ws示例的评论
df.withColumn("colmn", concat_ws("|!|", $"IsRangeAllowed_1", "IsRangeAllowed", ...)
.selectExpr("colmn")
.show()
//to add all columns in df
df.withColumn("colmn", concat_ws("|!|", df.cols:_*))
.selectExpr("colmn")
.show()