像Excel一樣使用SQL進行數據分析
- 2019 年 11 月 22 日
- 筆記
Excel是數據分析中最常用的工具 ，利用Excel可以完成數據清洗，預處理，以及最常見的數據分類，數據篩選，分類匯總，以及數據透視等操作，而這些操作用SQL一樣可以實現。SQL不僅可以從資料庫中讀取數據，還能通過不同的SQL函數語句直接返回所需要的結果，從而大大提高了自己在客戶端應用程式中計算的效率。
1 重複數據處理
- 查找重複記錄
SELECT * FROM user Where (nick_name,password) in ( SELECT nick_name,password FROM user group by nick_name,password having count(nick_name)>1 );
- 查找去重記錄
查找id最大的記錄
SELECT * FROM user WHERE id in (SELECT max(id) FROM user group by nick_name,password having count(nick_name)>1 );
- 刪除重複記錄
只保留id值最小的記錄
DELETE c1 FROM customer c1,customer c2 WHERE c1.cust_email=c2.cust_email AND c1.id>c2.id;
DELETE FROM user Where (nick_name,password) in (SELECT nick_name,password FROM (SELECT nick_name,password FROM user group by nick_name,password having count(nick_name)>1) as tmp1 ) and id not in (SELECT id FROM (SELECT min(id) id FROM user group by nick_name,password having count(nick_name)>1) as tmp2 );
2 缺失值處理
- 查找缺失值記錄
SELECT * FROM customer WHERE cust_email IS NULL;
- 更新列填充空值
UPDATE sale set city = "未知" WHERE city IS NULL; UPDATE orderitems set price_new=IFNULL(price_new,5.74);
- 查詢並填充空值列
SELECT AVG(price_new) FROM orderitems; SELECT IFNULL(price_new,5.74) AS bus_ifnull FROM orderitems;
3 計算列
- 更新表添加計算列
ALTER TABLE orderitems ADD price_new DECIMAL(8,2) NOT NULL; UPDATE orderitems set price_new= item_price*count;
- 查詢計算列
SELECT item_price*count as sales FROM orderitems;
4 排序
- 多列排序
SELECT * FROM orderitems ORDER BY price_new DESC,quantity;
- 查詢排名前幾的記錄
SELECT * FROM orderitems ORDER BY price_new DESC LIMIT 5;
- 查詢第10大的值
SELECT DISTINCT price_new FROM orderitems ORDER BY price_new DESC LIMIT 9,1;
- 排名
數值相同的排名相同且排名連續
SELECT prod_price, (SELECT COUNT(DISTINCT prod_price) FROM products WHERE prod_price>=a.prod_price ) AS rank FROM products AS a ORDER BY rank ;
5 字元串處理
- 字元串替換
UPDATE data1 SET city=REPLACE(city,'SH','shanghai'); SELECT city FROM data1;
- 按位置字元串截取
字元串截取可用於數據分列 MySQL 字元串截取函數：left(), right(), substring(), substring_index()
SELECT left('example.com', 3);
從字元串的第 4 個字元位置開始取，直到結束
SELECT substring('example.com', 4);
從字元串的第 4 個字元位置開始取，只取 2 個字元
SELECT substring('example.com', 4, 2);
- 按關鍵字截取字元串
取第一個分隔符之前的所有字元，結果是www
SELECT substring_index('www.google.com','.',1);
取倒數第二個分隔符之後的所有字元，結果是google.com;
SELECT substring_index('www.google.com','.',-2);
6 篩選
- 通過操作符實現高級篩選
使用 AND OR IN NOT 等操作符實現高級篩選過濾
SELECT prod_name，prod_price FROM Products WHERE vend_id IN('DLL01','BRS01'); SELECT prod_name FROM Products WHERE NOT vend_id='DLL01';
- 通配符篩選
常用通配符有% _ [] ^
SELECT * from customers WHERE country LIKE "CH%";
7 表聯結
SQL表連接可以實現類似於Excel中的Vlookup函數的功能
SELECT vend_id,prod_name,prod_price FROM Vendors INNER JOIN Products ON Vendors.vend_id=Products.vend_id; SELECT prod_name,vend_name,prod_price,quantity FROM OderItems,Products,Vendors WHERE Products.vend_id=Vendors.vend_id AND OrderItems.prod_id=Products.prod_id AND order_num=20007;
自聯結 在一條SELECT語句中多次使用相同的表
SELECT c1.cust_od,c1.cust_name,c1.cust_contact FROM Customers as c1,Customers as c2 WHERE c1.cust_name=c2.cust_name AND c2.cust_contact='Jim Jones';
8 數據透視
數據分組可以實現Excel中數據透視表的功能
- 數據分組
group by 用於數據分組 having 用於分組後數據的過濾
SELECT order_num,COUNT(*) as items FROM OrderItems GROUP BY order_num HAVING COUNT(*)>=3;
- 交叉表
通過CASE WHEN函數實現
SELECT data1.city, CASE WHEN colour = "A" THEN price END AS A, CASE WHEN colour = "B" THEN price END AS B, CASE WHEN colour = "C" THEN price END AS C, CASE WHEN colour = "F" THEN price END AS F FROM data1
