Scala创建DataFrame的几种方式

1 通过数据源创建DF

原始数据：

{"name":"Tom","age":18},
{"name":"Alice","age":17}

步骤：

//读取目录文件
scala> val df = spark.read.json("file:///opt/module/spark/mycode/a.json")
df: org.apache.spark.sql.DataFrame = [age: bigint, name: string]

//创建临时视图
scala> df.createOrReplaceTempView("people")

//查看DF
scala> spark.sql("select * from people").show

//打印结果
+---+-----+
|age| name|
+---+-----+
| 18|  Tom|
| 17|Alice|
+---+-----+

2 通过反射机制创建DF

原始数据：

Tom,18
Alice,17

步骤：

//导入包，支持把一个RDD隐式转换为一个DataFrame
scala> import spark.implicits._ 

//1.创建样例类
case class People(name:String, age: Int)

//2.创建RDD
scala> val rdd = sc.textFile("file:///opt/module/spark/mycode/b.txt")
rdd: org.apache.spark.rdd.RDD[String] = file:///opt/module/spark/mycode/b.txt MapPartitionsRDD[1] at textFile at :24

//3.使用map算子将每一行都创建一个People对象
scala> val df = rdd.map(_.split(",")).map(data=>People(data(0),data(1).toInt)).toDF
df: org.apache.spark.sql.DataFrame = [name: string, age: int]

//4.打印df
scala> df.show
+-----+---+
| name|age|
+-----+---+
|  Tom| 18|
|Alice| 17|
+-----+---+

注意，import spark.implicits._ 中的spark指的是SparkSession对象（在scala交互式环境中默认存在了一个SparkSession对象）。

启动spark-shell时的信息：

3 使用编程定义方式创建DF

当无法提前定义case class时，就需要采用编程方式定义RDD模式

//导入相关包
scala> import org.apache.spark.sql.types._
import org.apache.spark.sql.types._ 
scala> import org.apache.spark.sql.Row
import org.apache.spark.sql.Row 

//1.定义一个模式字符串
scala> val schemaString = "name age"
schemaString: String = name age

//2.根据模式字符串创建模式
scala> val fields = schemaString.split(" ").map(fieldName=>StructField(fieldName,StringType,nullable=true))
fields: Array[org.apache.spark.sql.types.StructField] = Array(StructField(name,StringType,true), StructField(age,StringType,true))

scala> val schema = StructType(fields)
schema: org.apache.spark.sql.types.StructType = StructType(StructField(name,StringType,true), StructField(age,StringType,true))

//3.将rdd的每条数据转换成Row类型
scala> val rowRDD = rdd.map(_.split(",")).map(data=>Row(data(0),data(1)))
rowRDD: org.apache.spark.rdd.RDD[org.apache.spark.sql.Row] = MapPartitionsRDD[19] at map at <console>:29

//4.根据设计的模式创建DF
scala> val df = spark.createDataFrame(rowRDD, schema)
df: org.apache.spark.sql.DataFrame = [name: string, age: string]

//5.打印
scala> df.show
+-----+---+                                                                     
| name|age|
+-----+---+
|  Tom| 18|
|Alice| 17|
+-----+---+