用Spark写UDF

操作的数据：

{"city":"St. John's","avgHigh":8.7,"avgLow":0.6}
{"city":"Charlottetown","avgHigh":9.7,"avgLow":0.9}
{"city":"Halifax","avgHigh":11.0,"avgLow":1.6}
{"city":"Fredericton","avgHigh":11.2,"avgLow":-0.5}
{"city":"Quebec","avgHigh":9.0,"avgLow":-1.0}
{"city":"Montreal","avgHigh":11.1,"avgLow":1.4}

object UdfDemo1 {
  def main(args: Array[String]): Unit = {
    val conf: SparkConf = new SparkConf().setAppName("").setMaster("local")
    val sc = new SparkContext(conf)
    // 可以执行SQL语句的实例
    val sct: SQLContext = new SQLContext(sc)
	// 以json格式读取文件
    val file: DataFrame = sct.read.json("E://words.txt")

    // 创建表
    file.createOrReplaceTempView("citytemps")
	// 注册一个函数，  输出的每个数都进行处理
    sct.udf.register("CTOF",(dc:Double)=>((dc*9.0/5.0)+32.0))
	// 调用函数  		自定义函数(参数)
    sct.sql("select city,CTOF(avgLow) as avgLows, CTOF(avgHigh) as avgHighs from citytemps").show()

    sc.stop()
  }
}

输出结果：

+-------------+-------+--------+
|         city|avgLows|avgHighs|
+-------------+-------+--------+
|         null|   null|    null|
|Charlottetown|  33.62|   49.46|
|      Halifax|  34.88|    51.8|
|  Fredericton|   31.1|   52.16|
|       Quebec|   30.2|    48.2|
|     Montreal|  34.52|   51.98|
+-------------+-------+--------+