clickhouse MPPDB数据库 实现复杂功能的SQL示例

环境:

sql 复制代码
clickhouse-client -m -h CH01--port 9001 -d test

伪MAC判断

sql 复制代码
CK通过SQL判断假MAC的示例:  select has(['2','6','A','E','a','e'],SUBSTRING('03-92-26-5B-13-02',2,1) )

将MAC地址替换为字段名即可

1.相似车分析

1.1执行

sql 复制代码
SELECT length(arrayFilter((x, y) -> (x != y), arrayMap(x -> substringUTF8(HPHM, x + 1, 1), range(7)), b)) AS diffLen,HPHM
	FROM
(SELECT PLATE_NO AS HPHM,['滇', '0', '1', 'A', '2', '2', '1'] AS b from CAR_DETECT_RL)
	WHERE diffLen < 2 limit 10;
	
	
SELECT length(arrayFilter((x, y) -> (x != y), arrayMap(x -> substringUTF8(HPHM, x + 1,1), range(7)), b)) AS diffLen,HPHM FROM (SELECT '滇04A221' AS HPHM,['滇', '0', '1', 'A', '2', '2', '1'] AS b from numbers(3)) WHERE diffLen < 2 limit 10;

1.2结果

tex 复制代码
┌─diffLen─┬─HPHM─────┐
│       1 │ 滇04A221 │
└─────────┴──────────┘
┌─diffLen─┬─HPHM─────┐
│       1 │ 滇C1A221 │
└─────────┴──────────┘
┌─diffLen─┬─HPHM─────┐
│       1 │ 滇01AD21 │
└─────────┴──────────┘

2.车辆轨迹搜车

2.1执行

sql 复制代码
--判断车辆实际运行轨迹 是否包含指定的运行轨迹
SELECT
	PLATE_NO,
    length(b) AS len2,
    indexOf(a, b[1]) AS idx,
    length(a) AS len1,
    if(len1 < (len2 + idx), 100, length(arrayFilter(x -> ((a[idx + x]) != (b[x + 1])), range(len2)))) AS diffLen
FROM
(
    select PLATE_NO,groupArray(DEVICE_ID) AS a,arrayDistinct(['4401000000131000301', '494583076246736505','4401000000131000320']) 
    AS b from (
    	SELECT PLATE_NO,DEVICE_ID
    	FROM CAR_DETECT_RL where PASS_TIME between '2019-07-01 00:00:00' and '2019-07-01 23:59:59' order by PASS_TIME
    ) t group by PLATE_NO
) where diffLen > 2;


SELECT PLATE_NO,
	length(b) AS len2,
	indexOf(a, b[1]) AS idx,
	length(a) AS len1,
	if(len1 < (len2 + idx), 100, length(arrayFilter(x -> ((a[idx + x]) != (b[x + 1])), range(len2)))) AS diffLen 
	FROM ( 
		select '浙4C12C2' as PLATE_NO,
		array('4401000000131000301', '494583076246736505','4401000000131000320') AS a,arrayDistinct(['4401000000131000301', '494583076246736505','4401000000131000320']
 	) AS b ) where diffLen > 2;

2.2结果

tex 复制代码
┌─PLATE_NO─┬─len2─┬─idx─┬─len1─┬─diffLen─┐
│ 浙4C12C2 │    3 │  11 │   15 │       1 │
│ 湘2D44DB │    3 │   3 │    9 │       1 │
│ 鲁B1BB1D │    3 │   4 │    7 │       1 │
│ 鲁B1CA10 │    3 │   2 │   13 │       1 │
│ 湘BAB2A2 │    3 │  11 │   16 │       1 │
│ 赣A022B2 │    3 │   4 │   10 │       1 │
│ 鲁D2B410 │    3 │   7 │   11 │       1 │
│ 云DD002C │    3 │  11 │   15 │       1 │

2.3 测试SQL

sql 复制代码
SELECT a,b,if(len1<len2+idx,100,length(arrayFilter(x-> (a[idx+x] != b[x+1]),range(len2)))) as diffLen  
FROM ( 
	SELECT  arrayDistinct(['A', 'B', 'C', 'D','D','D', 'H', 'F', 'G']) AS a,
	arrayDistinct(['B', 'C', 'D', 'E', 'F']) AS b,length(a) AS len1,length(b) AS len2,indexOf(a,b[1]) as idx) 
where diffLen<2;
相关推荐
不羁。。4 小时前
【撸靶笔记】第八关:GET - Blind - Boolian Based - Single Quotes
数据库·sql·mybatis
AwhiteV4 小时前
利用图数据库高效解决 Text2sql 任务中表结构复杂时占用过多大模型上下文的问题
数据库·人工智能·自然语言处理·oracle·大模型·text2sql
m0_595199855 小时前
Redis(以Django为例,含具体操作步骤)
数据库·redis·缓存
爱尚你19935 小时前
MySQL 三大日志:redo log、undo log、binlog 详解
数据库·mysql
小猿姐6 小时前
KubeBlocks AI:AI时代的云原生数据库运维探索
数据库·人工智能·云原生·kubeblocks
NocoBase8 小时前
10 个开源工具,快速构建数据应用
数据库·低代码·开源
麻辣清汤8 小时前
结合BI多维度异常分析(日期-> 商家/渠道->日期(商家/渠道))
数据库·python·sql·finebi
Kan先生10 小时前
对象存储解决方案:MinIO 的架构与代码实战
数据库·python
超级迅猛龙10 小时前
保姆级Debezium抽取SQL Server同步kafka
数据库·hadoop·mysql·sqlserver·kafka·linq·cdc
杨过过儿10 小时前
【Task02】:四步构建简单rag(第一章3节)
android·java·数据库