HDPCD-Java-复习笔记（21）- lab

Younge__

于 2017-10-21 10:02:18 发布

阅读量325

点赞数

分类专栏： hdp hadoop 文章标签： hdp

本文链接：https://blog.youkuaiyun.com/yongaini10/article/details/78301536

版权

hadoop 同时被 2 个专栏收录

24 篇文章

订阅专栏

hdp

23 篇文章

订阅专栏

本文介绍如何使用Hive User Defined Function (UDF) 来实现股票市场中资金流量指标的计算，并提供了从创建表、加载数据到定义及调用UDF的完整过程。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

Java lab booklet

Lab: Writing a Hive User Defined Function (UDF)

importstocks.hive

CREATE TABLE stocks (xchange STRING,
symbol STRING,
priceDate STRING,
open FLOAT,
high FLOAT,
low FLOAT,
close FLOAT,
volume INT,
adjClose FLOAT)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY ',';

LOAD DATA LOCAL INPATH '/root/java/labs/data/stock_prices/NYSE_daily_prices_A.csv' OVERWRITE INTO TABLE stocks;

MoneyFlow.java

package hiveudfs;

import org.apache.hadoop.hive.ql.exec.UDF;
import org.apache.hadoop.hive.serde2.io.DoubleWritable;

public class MoneyFlow extends UDF {
  private DoubleWritable result = new DoubleWritable();

  public DoubleWritable evaluate(double high, double low, double close, int volume) {
    double typicalPrice = (high + low + close) / 3;
    double moneyFlow = typicalPrice * volume;
    result.set(moneyFlow);
    return result;
  }
}

moneyflow.hive

ADD JAR hiveudfs.jar;
CREATE TEMPORARY FUNCTION moneyflow AS 'hiveudfs.MoneyFlow';
SELECT symbol, priceDate, moneyflow(high, low, close, volume) FROM stocks LIMIT 10000;