统计各个省份所属城市下最受欢迎的Top 3产品和其销量(不能出现有null)
所需数据表包括订单明细表和每日新增用户表
可以看出订单明细表存放的是产品信息,而每日新增用户表存放的是客户信息。
所需字段:城市,产品,销量
第一步:订单明细表和每日新增用户表连接,并求得每个城市下每个产品的订单量
SELECT a.chinese_city AS city ,b.product_key,COUNT(b.sales_order_key) AS sale_amount
FROM ods_customer a LEFT JOIN ods_sales_orders b ON a.customer_key = b.customer_key
GROUP BY a.chinese_city,b.product_key
第二步:对各城市下产品的订单量进行排序
注意:需要对城市做分组处理,并且去除订单量为0的记录,也就是该城市没有消费过该产品
SELECT c.city AS city,c.product_key AS product_key,c.sale_amount AS sale_amount,
row_number() over(PARTITION BY c.city ORDER BY c.sale_amount DESC )AS rank
FROM(
SELECT a.chinese_city AS city ,b.product_key,COUNT(b.sales_order_key) AS sale_amount
FROM ods_customer a LEFT JOIN ods_sales_orders b ON a.customer_key = b.customer_key
GROUP BY a.chinese_city,b.product_key
)c
WHERE c.sale_amount>0
第三步:选出各城市销量为Top3的产品
SELECT d.city AS city,d.product_key AS product_key,d.sale_amount AS sale_amount,d.rank AS rank
FROM(
SELECT c.city AS city,c.product_key AS product_key,c.sale_amount AS sale_amount,
row_number() over(PARTITION BY c.city ORDER BY c.sale_amount DESC )AS rank
FROM(
SELECT a.chinese_city AS city ,b.product_key,COUNT(b.sales_order_key) AS sale_amount
FROM ods_customer a LEFT JOIN ods_sales_orders b ON a.customer_key = b.customer_key
GROUP BY a.chinese_city,b.product_key
)c
WHERE c.sale_amount>0
)d
WHERE d.rank <=3;
结果局部截图