为什么mysql语句在SparkSql里报错?

Question

spark.sql("select qj,sd,cp from (select cp,sd,qj,row_number() over(partition by qj order by sd desc) rk from car) where rk>=1 and rk<=3").show()
    println("对比!!!!!!!!!")
    spark.sql("select * from car c where 3 > (select count(*) from car  where qj = c.qj and sd > c.sd) order by c.qj,c.sd desc").show()

第二句在mysql里能够正常运行,在Scala的SparkSql里运行报错

在mysql里效果是这样的

Answer 1

https://blog.csdn.net/chenyulancn/article/details/78682862