Real MySQL interview question (18) -- practical operation analysis

There is... In a company's database 3 A watch , Sales order form 、 Product list 、 List of sales outlets .

” Sales order form ” Recorded sales , Each piece of data indicates which customer 、 On which day 、 Which outlet bought what products , What is the number of purchases , And the retail price of the corresponding product .

“ Product list ” Records the details of the company's products .

“ List of sales outlets ” Recorded the company's sales outlets .

The sales order table and product detail table are passed through “ product ” Field Association , The sales order form and sales outlets are passed through “ Trading outlets ” relation .

#1. Analysis in 2020 Number of buyers in the first quarter of the year , Sales amount , Customer unit price , A single guest , Purchase frequency per capita 
SELECT 
  COUNT(DISTINCT  customer ID) AS  Number of buyers ,
  SUM( sales volumes  *  Retail price ) AS  Sales amount ,
  SUM( sales volumes  *  Retail price ) / COUNT(DISTINCT  customer ID) AS  Customer unit price ,
  SUM( sales volumes ) / COUNT(DISTINCT  customer ID) AS  A single guest ,
  COUNT(DISTINCT  The order number ) / COUNT(DISTINCT  customer ID) AS  Purchase frequency per capita  
FROM
   Sales order form  
WHERE  Transaction date  BETWEEN '2020-01-01' 
  AND '2020-03-31' 

#2. Analyze the brand in 2019.5-2020.4 Repurchase rate during the period （ Repeat purchase rate ： Place orders during this period >1 The number of people / Total number of buyers ）
SELECT 
  SUM(
    CASE
      WHEN  Order each customer  > 1 
      THEN 1 
      ELSE 0 
    END
  ) / COUNT(*) AS  Repeat purchase rate  
FROM
  (SELECT 
     customer ID,
    COUNT(DISTINCT  The order number ) AS  Order each customer  
  FROM
     Sales order form  
  WHERE  Transaction date  BETWEEN '2019-05-01' 
    AND '2020-04-30' 
  GROUP BY  customer ID) a 

#3. Find both purchased ProductA Bought again ProductB, But didn't buy ProductC Number of customers , And calculate the average customer unit price 
SELECT 
  COUNT(DISTINCT  customer ID) AS  Number of customers ,
  SUM( sales volumes  *  Retail price ) / COUNT(DISTINCT  The order number ) AS  Average customer price  
FROM
   Sales order form  
WHERE  customer ID IN 
  (SELECT 
     customer ID 
  FROM
    (SELECT 
       customer ID,
      GROUP_CONCAT(DISTINCT  product ) AS  Product portfolio  
    FROM
       Sales order form  
    GROUP BY  customer ID) a 
  WHERE  Product portfolio  LIKE '%ProductA%' 
    AND  Product portfolio  LIKE '%ProductB%' 
    AND  Product portfolio  NOT LIKE '%ProductC%') 
GROUP BY  customer ID 

Image Name

#4. Find the second largest customer in each city , List the cities they bought 、 customer ID、 Purchase amount 
SELECT 
  * 
FROM
  (SELECT 
    *,
    dense_rank () over (
      PARTITION BY  City  
  ORDER BY  Purchase amount  DESC
  ) AS  ranking  
  FROM
    (SELECT 
      b. City ,
      a. customer ID,
      SUM( sales volumes  *  Retail price ) AS  Purchase amount  
    FROM
       Sales order form  a 
      JOIN  List of sales outlets  b 
        ON a. Trading outlets  = b. Trading outlets  
    GROUP BY b. City ,
      a. customer ID) a) b 
WHERE  ranking  = 2

#5. Calculate the number of stores in each city and the business summary of each city , The output contains cities without purchase records 
SELECT 
   City ,
  COUNT(DISTINCT  Trading outlets ) AS  Number of stores in the city ,
  SUM( sales volumes  *  Retail price ) AS  Business summary  
FROM
   List of sales outlets  
  LEFT JOIN  Sales order form  USING ( Trading outlets ) 
GROUP BY  City