学习随笔 pyspark JDBC 操作oracle数据库

# -*- coding:utf-8 -*-
from pyspark import SparkContext, SparkConf
from pyspark.sql import SQLContext
import numpy as np


appName = "jhl_spark_1"  # 你的应用程序名称
master = "local"  # 设置单机
conf = SparkConf().setAppName(appName).setMaster(master)  # 配置SparkContext
sc = SparkContext(conf=conf)
sqlContext = SQLContext(sc)
url='jdbc:oracle:thin:@127.0.0.1:1521:ORCL'
tablename='V_JSJQZ'
properties={"user": "Xho", "password": "sys"}
df=sqlContext.read.jdbc(url=url,table=tablename,properties=properties)
#df=sqlContext.read.format("jdbc").option("url",url).option("dbtable",tablename).option("user","Xho").option("password","sys").load()
#注册为表，然后在SQL语句中使用
df.registerTempTable("v_jsjqz")
#SQL可以在已注册为表的RDDS上运行
df2=sqlContext.sql("select ZBLX,BS,JS,JG from v_jsjqz t order by ZBLX,BS")
list_data=df2.toPandas()# 转换格式toDataFrame
list_data = list_data.dropna()# 清洗操作，去除有空值的数据
list_data = np.array(list_data).tolist()#tolist
RDDv1=sc.parallelize(list_data)#并行化数据，转化为RDD
RDDv2=RDDv1.map(lambda x:(x[0]+'^'+x[1],[[float(x[2]),float(x[3])]]))
RDDv3=RDDv2.reduceByKey(lambda a,b:a+b)
sc.stop()

 这里的 pyspark 是spark安装的文件夹里python文件夹下的，需要复制到anoconda的Lib下site-packages中

代码中没有环境变量的配置，不愿意在本机配置环境变量的可以去查查spark在python中环境变量配置