当前位置：网站首页>Scala104-Spark.sql的内置日期时间函数
Scala104-Spark.sql的内置日期时间函数
2022-08-04 18:29:00 【51CTO】
有些时候我们会直接用df.createOrReplaceTempView(temp)创建临时表，用sql去计算。sparkSQL有些语法和hql不一样，做个笔记。
- <scala.version>2.11.12</scala.version>
- <spark.version>2.4.3</spark.version>
val
builder
=
SparkSession
.
builder()
.
appName(
"learningScala")
.
config(
"spark.executor.heartbeatInterval",
"60s")
.
config(
"spark.network.timeout",
"120s")
.
config(
"spark.serializer",
"org.apache.spark.serializer.KryoSerializer")
.
config(
"spark.kryoserializer.buffer.max",
"512m")
.
config(
"spark.dynamicAllocation.enabled",
false)
.
config(
"spark.sql.inMemoryColumnarStorage.compressed",
true)
.
config(
"spark.sql.inMemoryColumnarStorage.batchSize",
10000)
.
config(
"spark.sql.broadcastTimeout",
600)
.
config(
"spark.sql.autoBroadcastJoinThreshold",
-
1)
.
config(
"spark.sql.crossJoin.enabled",
true)
.
master(
"local[*]")
val
spark
=
builder.
getOrCreate()
spark.
sparkContext.
setLogLevel(
"ERROR")
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
builder: org.apache.spark.sql.SparkSession.Builder = [email protected]
spark: org.apache.spark.sql.SparkSession = [email protected]
- 1.
- 2.
var
df1
=
Seq(
(
1,
"2019-04-01 11:45:50",
11.15,
"2019-04-02 11:45:49"),
(
2,
"2019-05-02 11:56:50",
10.37,
"2019-05-02 11:56:51"),
(
3,
"2019-07-21 12:45:50",
12.11,
"2019-08-21 12:45:50"),
(
4,
"2019-08-01 12:40:50",
14.50,
"2020-08-03 12:40:50"),
(
5,
"2019-01-06 10:00:50",
16.39,
"2019-01-05 10:00:50")
).
toDF(
"id",
"startTimeStr",
"payamount",
"endTimeStr")
df1
=
df1.
withColumn(
"startTime",
$
"startTimeStr".
cast(
"Timestamp"))
.
withColumn(
"endTime",
$
"endTimeStr".
cast(
"Timestamp"))
df1.
printSchema
df1.
show()
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
root
|-- id: integer (nullable = false)
|-- startTimeStr: string (nullable = true)
|-- payamount: double (nullable = false)
|-- endTimeStr: string (nullable = true)
|-- startTime: timestamp (nullable = true)
|-- endTime: timestamp (nullable = true)
+---+-------------------+---------+-------------------+-------------------+-------------------+
| id| startTimeStr|payamount| endTimeStr| startTime| endTime|
+---+-------------------+---------+-------------------+-------------------+-------------------+
| 1|2019-04-01 11:45:50| 11.15|2019-04-02 11:45:49|2019-04-01 11:45:50|2019-04-02 11:45:49|
| 2|2019-05-02 11:56:50| 10.37|2019-05-02 11:56:51|2019-05-02 11:56:50|2019-05-02 11:56:51|
| 3|2019-07-21 12:45:50| 12.11|2019-08-21 12:45:50|2019-07-21 12:45:50|2019-08-21 12:45:50|
| 4|2019-08-01 12:40:50| 14.5|2020-08-03 12:40:50|2019-08-01 12:40:50|2020-08-03 12:40:50|
| 5|2019-01-06 10:00:50| 16.39|2019-01-05 10:00:50|2019-01-06 10:00:50|2019-01-05 10:00:50|
+---+-------------------+---------+-------------------+-------------------+-------------------+
df1: org.apache.spark.sql.DataFrame = [id: int, startTimeStr: string ... 4 more fields]
df1: org.apache.spark.sql.DataFrame = [id: int, startTimeStr: string ... 4 more fields]
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
timestamp转string
把timestamp转换成对应格式字符串
- date_format把timestamp转换成对应的字符串
- 字符串格式用"yyyyMMdd"表示
root
|-- yyyyMMdd: string (nullable = true)
|-- yyyy_MM_dd: string (nullable = true)
|-- yyyy: string (nullable = true)
+--------+----------+----+
|yyyyMMdd|yyyy_MM_dd|yyyy|
+--------+----------+----+
|20190401|2019-04-01|2019|
|20190502|2019-05-02|2019|
|20190721|2019-07-21|2019|
|20190801|2019-08-01|2019|
|20190106|2019-01-06|2019|
+--------+----------+----+
sql: String =
"
SELECT date_format(startTime,'yyyyMMdd') AS yyyyMMdd,
date_format(startTime,'yyyy-MM-dd') AS yyyy_MM_dd,
date_format(startTime,'yyyy') AS yyyy
FROM TEMP
"
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
timestamp转date
- to_date可以把timestamp转换成date类型
root
|-- startTime: timestamp (nullable = true)
|-- endTime: timestamp (nullable = true)
|-- startDate: date (nullable = true)
|-- endDate: date (nullable = true)
+-------------------+-------------------+----------+----------+
| startTime| endTime| startDate| endDate|
+-------------------+-------------------+----------+----------+
|2019-04-01 11:45:50|2019-04-02 11:45:49|2019-04-01|2019-04-02|
|2019-05-02 11:56:50|2019-05-02 11:56:51|2019-05-02|2019-05-02|
|2019-07-21 12:45:50|2019-08-21 12:45:50|2019-07-21|2019-08-21|
|2019-08-01 12:40:50|2020-08-03 12:40:50|2019-08-01|2020-08-03|
|2019-01-06 10:00:50|2019-01-05 10:00:50|2019-01-06|2019-01-05|
+-------------------+-------------------+----------+----------+
sql: String =
SELECT startTime,endTime,
to_date(startTime) AS startDate,
to_date(endTime) AS endDate
FROM TEMP
df2: org.apache.spark.sql.DataFrame = [startTime: timestamp, endTime: timestamp ... 2 more fields]
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
求时间差
- 天数差函数datediff可以应用在timestamp中，也可应用在date类型中，单位是自然天，而不是24小时
- 月份差函数months_between同样可以，月度的单位好像是不固定的，即31天or30天
df2.
createOrReplaceTempView(
"temp")
var
sql
=
"""
SELECT startTime,
endTime,
datediff(endTime,startTime) AS dayInterval1,
datediff(endDate,startDate) AS dayInterval2,
months_between(endTime,startTime) AS monthInterval1,
months_between(endDate,startDate) AS monthInterval2
FROM TEMP
"""
// spark.sql(sql).printSchema
spark.
sql(
sql).
show()
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
+-------------------+-------------------+------------+------------+--------------+--------------+
| startTime| endTime|dayInterval1|dayInterval2|monthInterval1|monthInterval2|
+-------------------+-------------------+------------+------------+--------------+--------------+
|2019-04-01 11:45:50|2019-04-02 11:45:49| 1| 1| 0.03225769| 0.03225806|
|2019-05-02 11:56:50|2019-05-02 11:56:51| 0| 0| 0.0| 0.0|
|2019-07-21 12:45:50|2019-08-21 12:45:50| 31| 31| 1.0| 1.0|
|2019-08-01 12:40:50|2020-08-03 12:40:50| 368| 368| 12.06451613| 12.06451613|
|2019-01-06 10:00:50|2019-01-05 10:00:50| -1| -1| -0.03225806| -0.03225806|
+-------------------+-------------------+------------+------------+--------------+--------------+
sql: String =
"
SELECT startTime,
endTime,
datediff(endTime,startTime) AS dayInterval1,
datediff(endDate,startDate) AS dayInterval2,
months_between(endTime,startTime) AS monthInterval1,
months_between(endDate,startDate) AS monthInterval2
FROM TEMP
"
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
Ref
2020-03-24 于南京市江宁区九龙湖
边栏推荐
- YOLOv7-Pose尝鲜，基于YOLOv7的关键点模型测评
- 敏捷开发项目管理的一些心得
- 数仓相关，总结
- leetcode/有效的回文串，含有不需要判断回文的字符
- Develop those things: How to obtain the traffic statistics of the monitoring site through the EasyCVR platform?
- 【注册荣耀开发者】赢【荣耀70】手机
- 火灾报警联网FC18中CAN光端机常见问题解答和使用指导
- Win10只读文件夹怎么删除
- After EasyCVR is locally connected to the national standard device to map the public network, the local device cannot play and cascade the solution
- 入选爱分析·银行数字化厂商全景报告，网易数帆助力金融数字化场景落地
猜你喜欢
袋鼠云思枢：数驹DTengine，助力企业构建高效的流批一体数据湖计算平台
EasyCVR calls the cloud recording API and returns an error and no recording file is generated. What is the reason?
关于使用腾讯云HiFlow场景连接器每天提醒签到打卡
Flask框架实现注册加密功能详解【Flask企业课学习】
八猴渲染器是什么？它能干什么？八猴软件的界面讲解
图解LeetCode——899. 有序队列（难度：困难）
如何进行自动化测试？
OpenInfra Days China 2022 | SelectDB to share with you the Apache Doris in Internet advertising business practices
当项目中自动格式化插件Prettier和ESLint冲突报错时如何解决
npm配置国内镜像（淘宝镜像）
随机推荐
MMDetection 使用示例：从入门到出门
How does EasyCVR call the double-speed playback of device recording through the interface?
How does the intelligent video surveillance platform EasyCVR use the interface to export iframe addresses in batches?
【STM32】入门（五）：串口TTL、RS232、RS485
limux入门3—磁盘与分区管理
Google Earth Engine APP——一键在线查看全球1984-至今年的影像同时加载一个影像分析
The Industrial Metaverse Brings Changes to Industry
Route lazy loading
基于 eBPF 的 Kubernetes 可观测实践
Google Earth Engine APP - one-click online viewing of global images from 1984 to this year and loading an image analysis at the same time
开篇-开启全新的.NET现代应用开发体验
Iptables防火墙基础知识介绍
EasyCVR如何通过接口调用设备录像的倍速回放？
如何封装 svg
FE01_OneHot-Scala Application
敏捷开发项目管理的一些心得
Kubernetes入门到精通- Operator 模式入门
dotnet core 使用 CoreRT 将程序编译为 Native 程序
链表的经典入门LeetCode题目
"No title"