文档反馈官招募中，报名立赚积分兑换代金券！> HOT
文档中心>腾讯云数据仓库 TCHouse-X>语法手册>SparkSQL 语法>DQL>SELECT>HAVING

HAVING

最近更新时间：2026-05-20 14:11:22

我的收藏

本页目录：

描述

HAVING 子句用于根据指定条件过滤 GROUP BY 产生的结果，通常与 GROUP BY 子句一起使用。

语法

HAVING boolean_expression

参数

子句/关键字
说明
boolean_expression
计算结果为布尔值的表达式。HAVING 中的表达式只能引用常量、GROUP BY 中的表达式或聚合函数

示例

-- 创建测试表
CREATE EXTERNAL TABLE hv_dealer (id INT, city STRING, car_model STRING, quantity INT)
    USING PARQUET LOCATION 'cosn://<your_cos_bucket>/test_having/hv_dealer';
INSERT INTO hv_dealer VALUES
    (100,'Fremont','Civic',10),(100,'Fremont','Accord',15),
    (200,'Dublin','Civic',20),(300,'San Jose','Accord',8);

-- HAVING column
SELECT city, sum(quantity) AS s FROM hv_dealer GROUP BY city HAVING city = 'Fremont';

-- HAVING aggregate
SELECT city, sum(quantity) AS s FROM hv_dealer GROUP BY city HAVING sum(quantity) > 15;

-- HAVING alias
SELECT city, sum(quantity) AS s FROM hv_dealer GROUP BY city HAVING s > 15;

-- HAVING different aggregate
SELECT city, sum(quantity) AS s FROM hv_dealer GROUP BY city HAVING max(quantity) > 15;

-- HAVING without GROUP BY
SELECT sum(quantity) AS s FROM hv_dealer HAVING sum(quantity) > 10;


中国站