Alibaba Tianchi SQL training camp task4 learning notes

1. Pay attention to distinguish the associated sub queries 、 The union of two tables 、 The table connects the relationship between the three .

2. Use NOT The predicate subtracts the set , Find out product In the table , The price is higher than 2000, But the profit is lower than 30% The goods , The results should be shown in the table below .

Solve this problem , use where＋and It can be fixed , But the problem here is to let you use not Predicate resolution , So in and The latter condition cannot be expressed directly , Use not in + The way of subquery , The syntax in this way is equivalent to a comparison conditional statement .

SELECT * FROM product WHERE sale_price > 2000 AND product_id NOT IN (SELECT product_id FROM product WHERE sale_price<1.3*purchase_price)

3. Use not in + union It can achieve the purpose of finding the symmetrical difference between two tables , But because in MySQL 8.0 in , Because the of two tables or query results cannot be obtained directly , Therefore, it is not suitable to use the above idea to calculate the symmetry difference . Fortunately, there are difference set operations that can be used . You can see intuitively , The difference in symmetry between two sets is equal to A-B And go up B-A, Therefore, in practice, we can use this idea to find the symmetry difference .

4. With the help of the Union and symmetric difference of the two tables, the purpose of finding the intersection can be achieved , The intersection of two sets can be regarded as the intersection of two sets, and the symmetry difference between the two sets is removed .

5. Writing order ：

SELECT》FROM 》WHERE》GROUP BY》HAVE》ORDER BY

Execution order ：

FROM 》WHERE》GROUP BY》HAVE》SELECT》ORDER BY

6. natural SQL The order of statement execution is to execute the subquery first , But the order of associated sub queries is to execute the main query first . The main purpose of this is to solve , The main query queries a row of records each time and the sub query directly outputs the aggregate query results . adopt “ relation ” To subquery , Limit the number of outputs of subqueries each time , So as to achieve the purpose of querying different kinds of aggregation results in a table , Therefore, the associated sub query needs to execute the main query first . It's kind of similar python in for Nesting of loops .

Associated subqueries and table joins can achieve the same effect , But why should there be table connections ？

If you have excel Of vlookup function , You will find that this function can also achieve this function . actually , In the way of thinking , Associated subqueries are more like vlookup function : By table A Main table , Then according to the table A The value of each row of the associated column , Go to the table one by one B Find rows with equal values in the associated columns in .
When the amount of data is small , There is no performance problem with this approach , But when the amount of data is large , This method will lead to large computational overhead : For each row of data returned by the external query , Will pass the value of an associated column to the internal sub query , Then the internal subquery executes a query based on the passed in value and returns its query result . This makes , For example, the returned result of the external main query has 10000 rows , Then the subquery will be executed 10000 times , This will lead to a terrible time consumption .

Why use associated subqueries ？

Sometimes you need to query the extension information in the same table , Do not find a subquery （ View ） As a reference, you cannot directly query , So at this time, you can use the associated sub query , Problems that can be queried with associated subqueries can also be solved with inner connections , Just embed the filter criteria in one of the connection tables in advance .

About within the link , The following three points need to be noted :

Key points : To make a connection, you need to be in FROM Use multiple tables in clause .

Previous FROM There is only one table in the clause , And this time we used shopproduct and product Two tables , Use keywords INNER JOIN You can link the two tables together :

FROMshopproduct AS SP INNER JOINproduct AS P

combination WHERE Clause uses inner links

The first increase WEHRE The way clause , The above query is used as a sub query , Enclosed in parentheses , Then add filter criteria to the outer query .

SELECT *
  FROM (--  Step 1: query results 
        SELECT SP.shop_id
               ,SP.shop_name
               ,SP.product_id
               ,P.product_name
               ,P.product_type
               ,P.sale_price
               ,SP.quantity
          FROMshopproduct AS SP
         INNER JOINproduct AS P
            ON SP.product_id = P.product_id) AS STEP1
 WHERE shop_name = ' Tokyo '
   AND product_type = ' clothes ' ;

Remember what we learned when we learned sub query ? The result of the subquery is actually a table , It's just a virtual table , It doesn't really exist in the database , Only other tables in the database are filtered , A result of query operations such as aggregation " View ".
This way of writing can clearly distinguish each operation step , We are not very familiar with SQL Querying the execution order of each clause can help us .

But actually , If we know WHERE Clause will be FROM Clause is followed by , in other words , Finishing INNER JOIN … ON Get a new table , Will execute WHERE Clause , Then you get the standard writing :

SELECT  SP.shop_id
       ,SP.shop_name
       ,SP.product_id
       ,P.product_name
       ,P.product_type
       ,P.sale_price
       ,SP.quantity
  FROMshopproduct AS SP
 INNER JOINproduct AS P
    ON SP.product_id = P.product_id
 WHERE SP.shop_name = ' Tokyo '
   AND P.product_type = ' clothes ' ;

（1） Write where Filter statements （2） After filtering in each query, connect the results

No matter how to solve the problem , It's important to think , Get a question , First analyze whether you need to connect 、 How to determine the screening criteria 、 Need grouping and merging , Consider these problems clearly before , Choose a suitable sentence to write .

Example ：

Find out the name and price of clothing goods in each store . The following results are expected :

The problem can be analyzed by sql The execution sequence of statements is substituted into the analysis step by step

1.from： Where are from ,product And shopproduct The two tables , Therefore, internal connection is required

shopproduct as sp inner join product as p on sp.product_id = p.product_id

2.where： What are the screening criteria , According to the meaning of the title, the condition is product_tape = ‘ clothes ’ 

3.select：SP.shop_id,SP.shop_name,SP.product_id ,P.product_name, P.product_type, P.purchase_price

（ For the sentences that do not need to be executed in the topic, you can directly skip ）

The final complete statement is ：

-- Refer to the answer 1-- Do not use subqueries

SELECT SP.shop_id,SP.shop_name,SP.product_id ,P.product_name, P.product_type, P.purchase_price FROM shopproduct AS SP INNER JOIN product AS P ON SP.product_id = P.product_id WHERE P.product_type = ' clothes ';

-- Refer to the answer 2-- Use subquery

SELECT SP.shop_id, SP.shop_name, SP.product_id ,P.product_name, P.product_type, P.purchase_price FROM shopproduct AS SP INNER JOIN -- from product Table to find out the information of clothing products (SELECT product_id, product_name, product_type, purchase_price FROM product WHERE product_type = ' clothes ')AS P ON SP.product_id = P.product_id;

The second method is to directly perform a subquery on the table used for connection , So the direct connection is the filtered sub query （ View ）, There is no need to filter after the connection .

Exercises :

In every store , What are the selling prices of the highest priced goods ?

--  Refer to the answer 
SELECT SP.shop_id
      ,SP.shop_name
      ,MAX(P.sale_price) AS max_price
  FROMshopproduct AS SP
 INNER JOINproduct AS P
    ON SP.product_id = P.product_id
 GROUP BY SP.shop_id,SP.shop_name

1.from

2.group by ,group by The next grouping field needs to be filled in according to the actual situation , It is not limited to finding a maximum value and grouping a field

3.select

Inner link and associated subquery

Find out the items in each category that sell at a price higher than the average price of the product .

Associated subquery

SELECT product_type, product_name, sale_price FROM product AS P1 WHERE sale_price > (SELECT AVG(sale_price) FROM product AS P2 WHERE P1.product_type = P2.product_type GROUP BY product_type);

Internal connection ：

SELECT P1.product_id ,P1.product_name ,P1.product_type ,P1.sale_price ,P2.avg_price FROM product AS P1 INNER JOIN (SELECT product_type,AVG(sale_price) AS avg_price FROM product GROUP BY product_type) AS P2 ON P1.product_type = P2.product_type WHERE P1.sale_price > P2.avg_price;

Ideas ：

1.from：product surface , There is no need to connect

2.where： The selling price of each product category sale_price Higher than the average selling price of this kind of goods avg_price Here is an aggregate function , So just use where Pick up group by Definitely not the right number , Therefore, you need to associate sub queries first , Extract the average selling price of each category , Then output the query results , It's a way of thinking .

In addition, we encounter such problems , Self connection can be considered to solve , Is the connection of the same table , Similar to the associated sub query of the same table . First step ： First, calculate the average selling price in groups ; The second step ： Make internal connection according to the type of goods ; The third step ： increase where Clause , Define the screening criteria .

summary ： For cases where records need to be compared with aggregate results , Or use sub queries to get aggregate results and compare them with each record ; or ： Aggregate first , Then make an internal connection with another table .

SELECT  P1.product_id
       ,P1.product_name
       ,P1.product_type
       ,P1.sale_price
       ,P2.avg_price
  FROM product AS P1
 INNER JOIN 
   (SELECT product_type,AVG(sale_price) AS avg_price 
      FROM product 
     GROUP BY product_type) AS P2 
    ON P1.product_type = P2.product_type
 WHERE P1.sale_price > P2.avg_price;

（task4-4.2.1.5） The background is ： Find out the items in each category that sell at a price higher than the average price of the product ？ Now we use inner connection to solve this problem , I want to ask you something ,P2 The table has passed group by Aggregated product_type, Why can it be combined with those that have not been aggregated P1 Table passing product_type Internal connection , Because “ON P1.product_type = P2.product_type”, I wrote this sentence , Even if the number is different, it can be connected ？

There is a mistake here ： Table connection does not require the same number of public fields , There is no need to have the same name , As long as the value of the upper number is right , So here we can directly aggregate the P2 With non aggregated P1 Connected .

Secondly, don't force yourself to find “ or ： Aggregate first , Then make an internal connection with another table .” This method is connected with the previous example “ Use subquery ” And “ Do not use subqueries ” There are two ways to find correlation , Because this question is to P1 The records in the table are consistent with P2 Records in which the average value is calculated by grouping aggregation are connected , It is fundamentally different from the two situations mentioned above , So there is no correlation between the two .

-------------------------------------------

This result is not consistent with the result given in the book , Less exercise T T-shirt , This is due to exercise T T-shirt regist_date Field is empty , When making natural connections , From product and product2 The movement of T When comparing this line of data , It's actually equivalent linking field by field , Remember when we were 6.2ISNULL,IS NOT NULL The comparison method of missing values learned in this section can be known , The two missing values are compared with an equal sign , The result is not true . The link will only return those lines that return true to the link condition .

Exercises : Use inner links to find product Table and product2 The intersection of tables .

SELECT P1.*
  FROMproduct AS P1
 INNER JOINproduct2 AS P2
    ON (P1.product_id  = P2.product_id
   AND P1.product_name = P2.product_name
   AND P1.product_type = P2.product_type
   AND P1.sale_price   = P2.sale_price
   AND P1.regist_date  = P2.regist_date)

The results are as follows

Note the above results and P230 The results are not consistent – Less product_id='0001' This business , Looking at the source table data, you can find , The less this line of data regist_date Is the missing value , Recall from Chapter 6 IS NULL The predicate , We learned that , This is because missing values cannot be compared with equal signs .

If we just use product_id To connect :

SELECT P1.*
  FROMproduct AS P1
 INNER JOINproduct2 AS P2
    ON P1.product_id = P2.product_id

Query results :

It's the same this time . Internal connections only connect common records .

The inner link will discard the unsatisfied in the two tables ON The conditions are right , The opposite of the inner link is the outer link . The outer link will selectively keep the unmatched rows according to the type of outer link .

According to which table the reserved rows are located , There are three forms of external links : Left link , Right link and all outer link .

The left link will be saved and cannot be followed in the left table ON Clause matches to the line , At this time, the rows corresponding to the right table are all missing values ; The right link will save the right table, which cannot be followed ON Clause matches to the line , At this time, the rows corresponding to the left table are all missing values ; The total external join will save two tables at the same time ON Clause matches to the line , The corresponding row in another table is filled with missing values .

The corresponding grammars of the three external links are :

--  Left link      
FROM <tb_1> LEFT  OUTER JOIN <tb_2> ON <condition(s)>
--  Right link      
FROM <tb_1> RIGHT OUTER JOIN <tb_2> ON <condition(s)>
--  All external links 
FROM <tb_1> FULL  OUTER JOIN <tb_2> ON <condition(s)>

Before, whether it is external connection or internal connection , A common prerequisite is the link condition –ON Clause , Used to specify the conditions of the link . If you have tried a link query without this link condition , You may have found out , There will be many lines . Remove... From the link ON Clause , It's called cross linking (CROSS JOIN), Cross linking is also called Cartesian product , The latter is a mathematical term . Two sets are Cartesian products , It's just using sets A Every element and set in B Each of the elements in makes up an ordered combination . Database table ( Or subquery ) And , Intersection and difference are operations such as expanding or filtering restrictions on the table vertically , This requires the number of columns in the table and the data type of the column in the corresponding position " Compatible with ", Therefore, these operations do not add new columns , And cross connect ( The cartesian product ) Is to expand the table horizontally , That is, add a new column , This is consistent with the function of the link . But because there is no ON Limitation of clause , Each row of the left table and the right table will be combined , This often leads to many meaningless rows appearing in the search results . Of course , In some query requirements , Cross linking also has some uses .

So in the associated sub query , Each piece of data in the main query will be combined with the sub query .

exercises ：

4.1select *
form product as p1 full outer join product2 as p2
where sale_price > 5000;

intersection ： The union of two sets removes the symmetry difference between the two sets

Symmetry difference ：A-B And go up B-A, Union of two difference sets

Difference set ：
A-B：select * from product where product_id not in  (select product_id from product 2);

B-A：select * from product2 where product_id not in  (select product_id from product 1);
Symmetry difference ：
select * from product where product_id not in  (select product_id from product 2)

union

select * from product2 where product_id not in  (select product_id from product 1);

4.2.select * from
(select * from product 

union 

select * from product2) as p1

where product_id not in 

(select * from product where product_id not in  (select product_id from product 2)

union

select * from product2 where product_id not in  (select product_id from product 1);

4.3
select p2.shop_name,p1.product_id,p1.product_type,max(sale_price)
from product as p1 right outer join shopproduct as p2 on product_id
group by product_type;

4.4
select product_type,product_id,sale_price
from product as p1 
where sale_price = (select product_type,product_id,max(sale_price)
                              from product as p2
                              where p1.product_type = p2.product_type
                              group by p2.product_type);

select product_type,product_id,sale_price
from product as p1 inner join (select product_type,product_id,max(sale_price)
                                              from product as p2
                                              group by p2.product_type)
                                              on p1.product_type = p2.product_type;

4.5
select product_id, produc_name, slae_price
from product
order by sale_price;