临汾市住房城乡建设局网站中国都有哪些网站

• 请编写一个Hive SQL查询，对于每个店铺，计算每一天的累计销售金额以及与前一天相比的销售金额增长额（如果是第一天，则增长额为0）。结果表应包含shop_id,sale_date,sale_amount,cumulative_sale_amount和sale_amount_increase这几列
  建表语句

CREATE TABLE sales_data (shop_id STRING,sale_date STRING,sale_amount DOUBLE
)
STORED AS PARQUET;  -- 可以选择不同的存储格式，如 Parquet, ORC, etc.
INSERT INTO sales_data VALUES('shop_1', '2023-01-01', 100.0),('shop_1', '2023-01-02', 150.0),('shop_1', '2023-01-03', 200.0),('shop_2', '2023-01-01', 80.0),('shop_2', '2023-01-02', 120.0),('shop_2', '2023-01-03', 160.0);

代码

select 
shop_id,
sale_date,
sale_amount,
sum(sale_amount) over(partition by shop_id order by sale_date) as cumulative_sale_amount,
if(lag(sale_amount) over(partition by shop_id order by sale_amount) is null , 0,sale_amount - lag(sale_amount) over(partition by shop_id order by sale_amount)
) as sale_amount_increase
from sales_data

结果