Ask 1 SQL Reading and Writing Problem 20

Simple. It's not challenging.

Ah, you have this book too, it's a good book, and the first textbook I used to teach SQL was this book.

I agree with the statement upstairs, first treat the next SQL statement as a table, and perform a joint query with the previous table!

--This should work, and the order of the results is not too different from yoursselect , count(*).

from test t1

join test t2 on = and order by ,

This is a bit difficult, my idea is that 2 functions, 1 stored procedure, 1 function, can find how many levels there are in this table.

1 function can return which is the root of a piece of data, which is the secondary root... Which is the nth layer root.

The stored procedure then organizes the statement according to the return value of the first function, sort by root, sort by secondary root. Nth layer root sorting.

select count(a.) as **Number from (select ****,**Price from table where date=3) a

join (select ****,**valence from table where date=2) b

on a.****=b.****。

where (a.Price-bPrice-bPrice between and

and left(a.****，1,5)='sh601'

aahuichao: how can you run without linking the two tables.

select count(a.****)

from (select ****,**price from the original table where **** like.)'sh601%'and date=3) a,(select ****,**price from the original table where **** like'sh601%'and date=2) b

where a.****=b.**** and (a.Price-b**Price) aPrice between and

Thank you downstairs for the correction.

The third floor is the right way, but the corresponding condition should not be the ID, but the Name.

dense_rank() over(partition by id order by id,time desc) ×

dense_rank() over(partition by name order by time desc) ○

If you want to extract the top 20, you can add a SQL statement in front of the top 20.

select top(20) id, name, time from (select id,name, time ,dense_rank() over(partition by name order by time desc) rank from table_name

x where rank <=3 order by time desc

Let's use the window function:

Don't know what database you're using:

The following SQL is available in both Db2 Oracle and SQL Server:

select * from (

select id,name, time ,dense_rank() over(partition by id order by id,time desc) dr

from table_name

xwhere dr <=3

Understand the dense rank function.

Added: select top 20 * from (select id,name, time ,dense rank() over(partition by id order by id,time desc) dr

from table_name

xwhere dr <=3

Upstairs is wrong, and what comes out is the first 3 records.

select top 3 *

from Staff Table.

group by name

order by time desc