Hive 我想从分区表中找出人口第二多的国家

我正在使用以下查询：

select country_name from (SELECT country_name, rank() OVER (ORDER BY TOTAL_pop desc) 
as rk FROM (SELECT country_name, sum(col8) as TOTAL_pop FROM world_bank_data_partitioned GROUP BY country_name)
) SUM_EMP where rk = 2;

并获取以下错误：

Error while compiling statement: FAILED: ParseException line 1:197 cannot recognize input near ')' 'SUM_EMP' 'where' in subquery source

---

请帮我做同样的事情。

您可以使用下面的SQL。您需要为每个子查询添加别名。我将子查询命名为

rs

select country_name from 
(SELECT country_name, rank() OVER (ORDER BY TOTAL_pop desc)  as rk FROM 
    (SELECT country_name, sum(col8) as TOTAL_pop FROM world_bank_data_partitioned GROUP BY country_name) rs 
) SUM_EMP where rk = 2;