Map读写性能

def main(args: Array[String]): Unit = {
    val t3 = System.currentTimeMillis()
    val key = "a"
    for (j <- 0 to 100000) {
      var expdata: scala.collection.mutable.Map[String, ArrayBuffer[Double]] = scala.collection.mutable.Map("a" -> ArrayBuffer[Double](1))
      for (i <- 0 to 10000) {
        if (expData.contains(key)) {
          expData(key) += i.toDouble
        } else {
          expData(key) = ArrayBuffer(i.toDouble)
        }
      }
    }
    val t4 = System.currentTimeMillis()
    println("expData 2 time : 执行" + (t4 - t3)/1000 + "秒" )

    val t5 = System.currentTimeMillis()
    val key2 = "a"
    for (j <- 0 to 100000) {
      var expdata: scala.collection.mutable.Map[String, ArrayBuffer[Double]] = scala.collection.mutable.Map("a" -> ArrayBuffer[Double](1))
      for (i <- 0 to 10000) {
        expData.put(key2,expData.getOrElse(key2, ArrayBuffer[Double]()) += i.toDouble)

      }
    }
    val t6 = System.currentTimeMillis()
    println("expData 3 time : 执行" + (t6 - t5)/1000 + "秒" )
  }

如图，更新map数据，分别使用两种方式（expData(key)、put|get）

1. expData(key) += i.toDouble

2. expData.put(key2,expData.getOrElse(key2, ArrayBuffer[Double]()) += i.toDouble)

处理了100000*10000=10亿次

结果如下：

expData(key)的执行32秒，put|get执行139秒

结论，expData(key)的速度是put|get的4倍多