对比SQL查询语句与Pandas语法（SQL vs Pandas）---基础篇

某些SQL查询语句与Pandas语句可以互相转换。

以下表（命名为df）为例：

      sex   tip  total_bill
0  Female  1.01       16.99
1    Male  1.66       10.34
2    Male  3.50       23.68
3    Male  3.31       23.68
4  Female  3.61       24.59

import pandas as pd
import numpy as np

df = pd.DataFrame({'total_bill': [16.99, 10.34, 23.68, 23.68, 24.59],
                   'tip': [1.01, 1.66, 3.50, 3.31, 3.61],
                   'sex': ['Female', 'Male', 'Male', 'Male', 'Female']})

1，选择

SQL：SELECT ... FROM ...

SELECT total_bill, tip 
FROM df

Pandas：loc， iloc

df.loc[:, ['total_bill', 'tip']]

2，条件过滤

SQL：WHERE

SELECT * 
FROM df 
WHERE total_bill > 20

Pandas：df[df[colunm] boolean expr]，.query()

df[df['total_bill'] > 20]

df.query('total_bill > 20')

3，限制行数

SQL：LIMIT

SELECT total_bill, tip 
FROM df 
LIMIT 3

Pandas：loc， iloc，.head()

df.loc[:2, ['total_bill', 'tip']]

df[['total_bill', 'tip']].head(3)

4，跳过指定行数

SQL：OFFSET

SELECT total_bill, tip 
FROM df 
LIMIT 3 OFFSET 1

Pandas：loc， iloc，.tail()

df.loc[1:3, ['total_bill', 'tip']]

df.loc[:3, ['total_bill', 'tip']].tail(3)

5，不重复的值

SQL：DISTINCT

SELECT DISTINCT sex 
FROM df 

Python：.unique()

df['sex'].unique()

6，条件连接

SQL：AND，OR

SELECT * 
FROM df
WHERE sex = "Female" 
    AND tip > 2

Pandas：&，|

df[(df['sex']=='Female') & (df['tip']>2)]

7，排序

SQL：ORDER BY

SELECT * 
FROM df 
ORDER BY total_bill DESC, tip

Pandas：.sort_values()

df.sort_values(['total_bill', 'tip'], ascending=[False, True])

8，在/不在...里面

SQL：IN，NOT IN

SELECT * 
FROM df 
WHERE tip NOT IN (1.66, 3.31)

Pandas：.isin()，~

df[~df['tip'].isin([1.66, 3.31])]

9，是否是空值

SQL：IS NULL，IS NOT NULL

SELECT total_bill 
FROM df 
WHERE total_bill IS NOT NULL

Pandas：.isnull()，.notnull()

df.total_bill[df['total_bill'].notnull()]

10，分组

SQL：GROUP BY

SELECT sex, COUNT(tip) 
FROM df 
GROUP BY sex

Pandas：.groupby()

df.groupby('sex')['tip'].count()

11，计数

SQL：COUNT

SELECT sex, COUNT(tip) 
FROM df 
GROUP BY sex

Pandas：.count()， .size()

df.groupby('sex')['tip'].count()

注：.count()返回非空字符出现的次数，相当于SQL中的COUNT(col_name)；而.size()返回所有行数，相当于SQL中的COUNT(*)。

12，对分组结果进行条件过滤

SQL：HAVING

SELECT sex, COUNT(tip) 
FROM df 
GROUP BY sex 
HAVING COUNT(tip) > 2

Pandas：

temp=df.groupby("sex")["tip"].count()
temp[temp>2]

13，聚合函数

SQL： MIN，MAX，AVG，SUM

SELECT sex, MAX(tip), SUM(total_bill) 
FROM df 
GROUP BY sex

Pandas：.agg()

df.groupby('sex').agg({'tip': np.max, 'total_bill': np.sum})

14，表连接

SQL：JOIN，LEFT JOIN，RIGHT JOIN

Pandas：.merge()

15， 表并集

SQL：UNION，UNION ALL

Pandas：pd.concat() + .drop_dupplicates()，pd.concat()