pandas与sql对应关系【帮助sql使用者快速上手pandas】

阿桨2025-01-12 23:36

本页旨在提供一些如何使用pandas执行各种SQL操作的示例，来帮助SQL使用者快速上手使用pandas。

目录

SQL语法

  • SELECT [DISTINCT | ALL] column1, column2, ..., aggregate_function(columnN), ...
  • FROM
  • table_name [AS alias]
  • [JOIN type JOIN table2_name [AS alias2] ON join_condition]
  • [, JOIN type JOIN table3_name [AS alias3] ON join_condition, ...]
  • [WHERE condition]
  • [GROUP BY column1, column2, ...]
  • [HAVING condition]
  • [ORDER BY column1 [ASC | DESC], column2 [ASC | DESC], ...]
  • [LIMIT number [OFFSET offset]]
  • [UNION [ALL] SELECT ...] -- 可以链式添加多个UNION SELECT语句
  1. DISTINCT：确保结果集中的行是唯一的。ALL（默认）表示返回所有匹配的行，包括重复的行。
  2. aggregate_function()：聚合函数，如**SUM(), AVG(), COUNT(), MAX(), MIN()**等，用于对一组值执行计算并返回单个值。
  3. JOIN type ：指定连接类型，如INNER JOIN, LEFT JOIN, RIGHT JOIN, FULL JOIN 等。ON join_condition：定义连接条件。
  4. WHERE condition：过滤结果集中的行，只返回满足条件的行。
  5. GROUP BY：将结果集按一个或多个列分组。通常与聚合函数一起使用。
  6. HAVING condition：过滤分组后的结果集，只返回满足条件的组。
  7. ORDER BY ：对结果集进行排序。可以指定多个列和排序方向（ASC 升序[默认]或DESC降序）。
  8. LIMIT number [OFFSET offset]：限制返回的行数，并可选地指定跳过的行数。
  9. UNION [ALL] ：合并两个或多个SELECT语句的结果集。UNION 默认去除重复行，而UNION ALL保留所有行。

一、选择SELECT

在SQL中，选择是使用要选择的列的逗号分隔列表（或* 选择所有列）

1、选择

SQL语法：

复制代码 
SELECT total_bill, tip, smoker, time
FROM data;

对应pandas实现：

复制代码 
In :data[["total_bill", "tip", "smoker", "time"]]
Out :
total_bill	tip	smoker	time
0	16.99	1.01	No	Dinner
1	10.34	1.66	No	Dinner
2	21.01	3.50	No	Dinner
3	23.68	3.31	No	Dinner
4	24.59	3.61	No	Dinner
...	...	...	...	...
239	29.03	5.92	No	Dinner
240	27.18	2.00	Yes	Dinner
241	22.67	2.00	Yes	Dinner
242	17.82	1.75	No	Dinner
243	18.78	3.00	No	Dinner
2、添加计算列

SQL语法：

复制代码 
SELECT *, tip/total_bill as tip_rate
FROM data;

对应pandas实现：

1）可以使用DataFrame的DataFrame.assign（）方法来追加新列

复制代码 
In :data = data.assign(tip_rate=data["tip"] / data["total_bill"])
In :data

Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808
...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744

2）也可以直接计算

复制代码 
In :data['tip_rate2'] = data["tip"] / data["total_bill"]
In :data

Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808
...	...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584	0.073584
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222	0.088222
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204	0.098204
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744	0.159744

二、连接JOIN ON

构造测试数据

复制代码 
In :df1 = pd.DataFrame({"key": ["A", "B", "C", "D"], "value": np.random.randn(4)})
In :df2 = pd.DataFrame({"key": ["B", "D", "D", "E"], "value": np.random.randn(4)})
1、内连接

SQL语法：

复制代码 
SELECT *
FROM df1
INNER JOIN df2
  ON df1.key = df2.key;

对应pandas实现：

复制代码 
In :pd.merge(df1, df2, on="key")
Out :	
key	value_x	value_y
0	B	0.227232	1.011278
1	D	1.415853	-0.149207
2	D	1.415853	-0.608430
2、左外连接

SQL语法：

复制代码 
SELECT *
FROM df1
LEFT OUTER JOIN df2
  ON df1.key = df2.key;

对应pandas实现：

复制代码 
In :pd.merge(df1, df2, on="key", how="left")
Out :	
key	value_x	value_y
0	A	1.418532	NaN
1	B	0.227232	1.011278
2	C	-0.578408	NaN
3	D	1.415853	-0.149207
4	D	1.415853	-0.608430
3、右外连接

SQL语法：

复制代码 
SELECT *
FROM df1
RIGHT OUTER JOIN df2
  ON df1.key = df2.key;

对应pandas实现：

复制代码 
In :pd.merge(df1, df2, on="key", how="right")
Out :
key	value_x	value_y
0	B	0.227232	1.011278
1	D	1.415853	-0.149207
2	D	1.415853	-0.608430
3	E	NaN	1.437388
4、全外连接

SQL语法：

复制代码 
SELECT *
FROM df1
FULL OUTER JOIN df2
  ON df1.key = df2.key;

对应pandas实现：

复制代码 
In :pd.merge(df1, df2, on="key", how="outer")
Out :
	key	value_x	value_y
0	A	1.418532	NaN
1	B	0.227232	1.011278
2	C	-0.578408	NaN
3	D	1.415853	-0.149207
4	D	1.415853	-0.608430
5	E	NaN	1.437388

三、过滤WHERE

SQL中的过滤是通过WHERE子句完成的。

SQL语法：

复制代码 
SELECT *
FROM data
WHERE total_bill >10;

对应pandas实现：

复制代码 
In :data[data["total_bill"] > 10]
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808
...	...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584	0.073584
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222	0.088222
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204	0.098204
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744	0.159744
1、AND

对应pandas中的&

SQL语法：

复制代码 
# 查询晚餐小费超过5美元的数据
SELECT *
FROM data
WHERE time = 'Dinner' AND tip > 5.00;

对应pandas实现：

复制代码 
In :data[(data["time"] == "Dinner") & (data["tip"] > 5.00)]
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
23	39.42	7.58	Male	No	Sat	Dinner	4	0.192288	0.192288
44	30.40	5.60	Male	No	Sun	Dinner	4	0.184211	0.184211
47	32.40	6.00	Male	No	Sun	Dinner	4	0.185185	0.185185
52	34.81	5.20	Female	No	Sun	Dinner	4	0.149382	0.149382
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
116	29.93	5.07	Male	No	Sun	Dinner	4	0.169395	0.169395
155	29.85	5.14	Female	No	Sun	Dinner	5	0.172194	0.172194
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812
172	7.25	5.15	Male	Yes	Sun	Dinner	2	0.710345	0.710345
181	23.33	5.65	Male	Yes	Sun	Dinner	2	0.242177	0.242177
183	23.17	6.50	Male	Yes	Sun	Dinner	4	0.280535	0.280535
211	25.89	5.16	Male	Yes	Sat	Dinner	4	0.199305	0.199305
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
214	28.17	6.50	Female	Yes	Sat	Dinner	3	0.230742	0.230742
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
2、OR

对应pandas中的|

SQL语法：

复制代码 
# 查询至少5名用餐者的小费或账单总额超过45美元的数据
SELECT *
FROM data
WHERE size >= 5 OR total_bill > 45;

对应pandas实现：

复制代码 
In :data[(data["size"] >= 5) | (data["total_bill"] > 45)]
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
125	29.80	4.20	Female	No	Thur	Lunch	6	0.140940	0.140940
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
142	41.19	5.00	Male	No	Thur	Lunch	5	0.121389	0.121389
143	27.05	5.00	Female	No	Thur	Lunch	6	0.184843	0.184843
155	29.85	5.14	Female	No	Sun	Dinner	5	0.172194	0.172194
156	48.17	5.00	Male	No	Sun	Dinner	6	0.103799	0.103799
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812
182	45.35	3.50	Male	Yes	Sun	Dinner	3	0.077178	0.077178
185	20.69	5.00	Male	No	Sun	Dinner	5	0.241663	0.241663
187	30.46	2.00	Male	Yes	Sun	Dinner	5	0.065660	0.065660
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
216	28.15	3.00	Male	Yes	Sat	Dinner	5	0.106572	0.106572
3、IS NULL

构造测试数据

复制代码 
In :frame = pd.DataFrame(
    {"col1": ["A", "B", np.nan, "C", "D"], "col2": ["F", np.nan, "G", "H", "I"]}
)

SQL语法：

复制代码 
SELECT *
FROM frame
WHERE col2 IS NULL;

对应pandas实现：

复制代码 
In :frame[frame["col2"].isna()]
Out :
col1	col2
1	B	NaN
4、IS NOT NULL

SQL语法：

复制代码 
SELECT *
FROM frame
WHERE col1 IS NOT NULL;

对应pandas实现：

复制代码 
In :frame[frame["col1"].notna()]
Out :
col1	col2
0	A	F
1	B	NaN
3	C	H
4	D	I
5、BETWEEN

SQL语法：

复制代码 
SELECT *
FROM data
WHERE tip between 5 and 7;

对应pandas实现：

复制代码 
In :data[data['tip'].between(5, 7)]
Out :
	total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
11	35.26	5.00	Female	No	Sun	Dinner	4	0.141804	0.141804
39	31.27	5.00	Male	No	Sat	Dinner	3	0.159898	0.159898
44	30.40	5.60	Male	No	Sun	Dinner	4	0.184211	0.184211
46	22.23	5.00	Male	No	Sun	Dinner	2	0.224921	0.224921
47	32.40	6.00	Male	No	Sun	Dinner	4	0.185185	0.185185
52	34.81	5.20	Female	No	Sun	Dinner	4	0.149382	0.149382
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
73	25.28	5.00	Female	Yes	Sat	Dinner	2	0.197785	0.197785
83	32.68	5.00	Male	Yes	Thur	Lunch	2	0.152999	0.152999
85	34.83	5.17	Female	No	Thur	Lunch	4	0.148435	0.148435
88	24.71	5.85	Male	No	Thur	Lunch	2	0.236746	0.236746
116	29.93	5.07	Male	No	Sun	Dinner	4	0.169395	0.169395
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
142	41.19	5.00	Male	No	Thur	Lunch	5	0.121389	0.121389
143	27.05	5.00	Female	No	Thur	Lunch	6	0.184843	0.184843
155	29.85	5.14	Female	No	Sun	Dinner	5	0.172194	0.172194
156	48.17	5.00	Male	No	Sun	Dinner	6	0.103799	0.103799
172	7.25	5.15	Male	Yes	Sun	Dinner	2	0.710345	0.710345
181	23.33	5.65	Male	Yes	Sun	Dinner	2	0.242177	0.242177
183	23.17	6.50	Male	Yes	Sun	Dinner	4	0.280535	0.280535
185	20.69	5.00	Male	No	Sun	Dinner	5	0.241663	0.241663
197	43.11	5.00	Female	Yes	Thur	Lunch	4	0.115982	0.115982
211	25.89	5.16	Male	Yes	Sat	Dinner	4	0.199305	0.199305
214	28.17	6.50	Female	Yes	Sat	Dinner	3	0.230742	0.230742
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
6、LIKE

开头/结尾字符匹配可以用startswith()/endswith()函数实现

SQL语法：

复制代码 
SELECT *
FROM data
WHERE time like 'Di%';

对应pandas实现：

复制代码 
In :data[data['time'].str.startswith('Di')]
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808
...	...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584	0.073584
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222	0.088222
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204	0.098204
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744	0.159744

中间字符匹配可以用contains()函数实现，na参数设置为False表示在缺失值上不返回True，case参数设置为False表示不区分大小写匹配

SQL语法：

复制代码 
SELECT *
FROM data
WHERE time like '%inne%';

对应pandas实现：

复制代码 
In :data[data['time'].str.contains('inne', na=False, case=False)]
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808
...	...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584	0.073584
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222	0.088222
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204	0.098204
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744	0.159744
7、CASE WHEN

SQL语法：

复制代码 
SELECT tip,case when tip<2 then 'LOW'
                when 2<=tip<=3 then 'MID'
                when 3<tip then 'HIG'
                end flag
FROM data;

对应pandas实现：

复制代码 
In :data['flag'] = data['tip'].apply(lambda x: 'LOW' if x < 2 else ('MID' if 2 <= x <= 3 else 'HIG'))
In :data[['tip', 'flag']]
Out :
	tip	flag
0	1.01	LOW
1	1.66	LOW
2	3.50	HIG
3	3.31	HIG
4	3.61	HIG
...	...	...
239	5.92	HIG
240	2.00	MID
241	2.00	MID
242	1.75	LOW
243	3.00	MID

四、分组GROUP BY

在pandas中，SQL的GROUP BY操作是使用类似名称的 groupby()方法。配合aggregate_function()使用

1、count()

SQL语法：

复制代码 
SELECT sex, count(*)
FROM data
GROUP BY sex;

对应pandas实现：

复制代码 
In :data.groupby("sex").size()
Out :
sex
Female     87
Male      157
dtype: int64
2、avg()

SQL语法：

复制代码 
SELECT day, AVG(tip), COUNT(*)
FROM tips
GROUP BY day;

对应pandas实现：

复制代码 
In :data.groupby("day").agg({"tip": "mean", "day": "size"})
Out :
tip	day
day		
Fri	2.734737	19
Sat	2.993103	87
Sun	3.255132	76
Thur	2.771452	62
3、sum()、max()、min()

SQL语法：

复制代码 
SELECT day, AVG(tip), SUM(tip), MAX(tip), MIN(tip), COUNT(tip)
FROM data
GROUP BY day;

对应pandas实现：

复制代码 
In :data.groupby("day").agg({
    "tip": ["mean", "sum", "max", "min"],
    "day": "size"
}).reset_index()
Out :
day	tip	day
mean	sum	max	min	size
0	Fri	2.734737	51.96	4.73	1.00	19
1	Sat	2.993103	260.40	10.00	1.00	87
2	Sun	3.255132	247.39	6.50	1.01	76
3	Thur	2.771452	171.83	6.70	1.25	62

五、HAVING

SQL语法：

复制代码 
SELECT day, AVG(tip), SUM(tip), MAX(tip), MIN(tip), COUNT(*)
FROM data
GROUP BY day
HAVING SUM(tip) > 200;

对应pandas实现：

复制代码 
In :result = data.groupby("day").agg({
    "tip": ["mean", "sum", "max", "min"],
    "day": "size"
}).reset_index()
In :result.columns = ['day', 'avg_tip', 'sum_tip', 'max_tip', 'min_tip', 'count_tips']
In :result[result['sum_tip'] > 200].reset_index()
Out :
	index	day	avg_tip	sum_tip	max_tip	min_tip	count_tips
0	1	Sat	2.993103	260.40	10.0	1.00	87
1	2	Sun	3.255132	247.39	6.5	1.01	76

六、排序ORDER BY

SQL语法：

复制代码 
SELECT *
FROM data
ORDER BY tip;

对应pandas实现：

复制代码 
In :data.sort_values("tip")
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
67	3.07	1.00	Female	Yes	Sat	Dinner	1	0.325733	0.325733
236	12.60	1.00	Male	Yes	Sat	Dinner	2	0.079365	0.079365
92	5.75	1.00	Female	Yes	Fri	Dinner	2	0.173913	0.173913
111	7.25	1.00	Female	No	Sat	Dinner	1	0.137931	0.137931
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
...	...	...	...	...	...	...	...	...	...
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
23	39.42	7.58	Male	No	Sat	Dinner	4	0.192288	0.192288
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812

SQL语法：

复制代码 
SELECT *
FROM data
ORDER BY tip,total_bill;

对应pandas实现：

复制代码 
In :data.sort_values(["tip","total_bill"])
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
67	3.07	1.00	Female	Yes	Sat	Dinner	1	0.325733	0.325733
92	5.75	1.00	Female	Yes	Fri	Dinner	2	0.173913	0.173913
111	7.25	1.00	Female	No	Sat	Dinner	1	0.137931	0.137931
236	12.60	1.00	Male	Yes	Sat	Dinner	2	0.079365	0.079365
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
...	...	...	...	...	...	...	...	...	...
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
23	39.42	7.58	Male	No	Sat	Dinner	4	0.192288	0.192288
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812

SQL语法：

复制代码 
SELECT *
FROM data
ORDER BY tip asc,total_bill desc;

对应pandas实现：

复制代码 
In :data.sort_values(by=["tip", "total_bill"], ascending=[True, False])
Out :
	total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
236	12.60	1.00	Male	Yes	Sat	Dinner	2	0.079365	0.079365
111	7.25	1.00	Female	No	Sat	Dinner	1	0.137931	0.137931
92	5.75	1.00	Female	Yes	Fri	Dinner	2	0.173913	0.173913
67	3.07	1.00	Female	Yes	Sat	Dinner	1	0.325733	0.325733
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
...	...	...	...	...	...	...	...	...	...
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
23	39.42	7.58	Male	No	Sat	Dinner	4	0.192288	0.192288
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812

七、LIMIT/OFFSET

1、LIMIT

在pandas中使用head()实现

SQL语法：

复制代码 
SELECT * 
FROM data
LIMIT 10;

对应pandas实现：

复制代码 
In :data.head(10)
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808
5	25.29	4.71	Male	No	Sun	Dinner	4	0.186240	0.186240
6	8.77	2.00	Male	No	Sun	Dinner	2	0.228050	0.228050
7	26.88	3.12	Male	No	Sun	Dinner	4	0.116071	0.116071
8	15.04	1.96	Male	No	Sun	Dinner	2	0.130319	0.130319
9	14.78	3.23	Male	No	Sun	Dinner	2	0.218539	0.218539
2、指定列中最大的前N行

SQL语法：

复制代码 
SELECT * 
FROM data
ORDER BY tip DESC
LIMIT 10;

对应pandas实现：

复制代码 
In :data.nlargest(10, columns="tip")
或
In :data.sort_values(by="tip", ascending=False).head(10)
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220
23	39.42	7.58	Male	No	Sat	Dinner	4	0.192288	0.192288
59	48.27	6.73	Male	No	Sat	Dinner	4	0.139424	0.139424
141	34.30	6.70	Male	No	Thur	Lunch	6	0.195335	0.195335
183	23.17	6.50	Male	Yes	Sun	Dinner	4	0.280535	0.280535
214	28.17	6.50	Female	Yes	Sat	Dinner	3	0.230742	0.230742
47	32.40	6.00	Male	No	Sun	Dinner	4	0.185185	0.185185
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
88	24.71	5.85	Male	No	Thur	Lunch	2	0.236746	0.236746
3、OFFSET

跳过排序后的前5行，选出接下来的10行

SQL语法：

复制代码 
SELECT * FROM tips
ORDER BY tip DESC
LIMIT 10 OFFSET 5;

对应pandas实现：

复制代码 
In :data.sort_values(by="tip", ascending=False).iloc[5:15]
Out :	
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2
214	28.17	6.50	Female	Yes	Sat	Dinner	3	0.230742	0.230742
183	23.17	6.50	Male	Yes	Sun	Dinner	4	0.280535	0.280535
47	32.40	6.00	Male	No	Sun	Dinner	4	0.185185	0.185185
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927
88	24.71	5.85	Male	No	Thur	Lunch	2	0.236746	0.236746
181	23.33	5.65	Male	Yes	Sun	Dinner	2	0.242177	0.242177
44	30.40	5.60	Male	No	Sun	Dinner	4	0.184211	0.184211
52	34.81	5.20	Female	No	Sun	Dinner	4	0.149382	0.149382
85	34.83	5.17	Female	No	Thur	Lunch	4	0.148435	0.148435
211	25.89	5.16	Male	Yes	Sat	Dinner	4	0.199305	0.199305

八、UNION ALL/UNION

pandas中使用concat()函数实现

构造测试数据

复制代码 
In :df1 = pd.DataFrame(
    {"city": ["Chicago", "San Francisco", "New York City"], "rank": range(1, 4)}
)
In :df2 = pd.DataFrame(
    {"city": ["Chicago", "Boston", "Los Angeles"], "rank": [1, 4, 5]}
)
1、UNION ALL

SQL语法：

复制代码 
SELECT city, rank
FROM df1
UNION ALL
SELECT city, rank
FROM df2;

对应pandas实现：

复制代码 
In :pd.concat([df1, df2])
Out :
city	rank
0	Chicago	1
1	San Francisco	2
2	New York City	3
0	Chicago	1
1	Boston	4
2	Los Angeles	5
2、UNION

SQL语法：

复制代码 
SELECT city, rank
FROM df1
UNION
SELECT city, rank
FROM df2;

对应pandas实现：

复制代码 
In :pd.concat([df1, df2]).drop_duplicates()
Out :
	city	rank
0	Chicago	1
1	San Francisco	2
2	New York City	3
1	Boston	4
2	Los Angeles	5

九、开窗函数

1、ROW_NUMBER()

为结果集中的每一行分配一个唯一的数字,顺序为1，2，3，4，5......

SQL语法：

查询每天total_bill最大的两行数据

复制代码 
SELECT * FROM (
  SELECT
    t.*,
    ROW_NUMBER() OVER(PARTITION BY day ORDER BY total_bill DESC) AS rn
  FROM data t
)
WHERE rn < 3
ORDER BY day, rn;

对应pandas实现：

复制代码 
In :(
    data.assign(
        rn=data.sort_values(["total_bill"], ascending=False)
        .groupby(["day"])
        .cumcount()
        + 1
    )
    .query("rn < 3")
    .sort_values(["day", "rn"])
)
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2	rn
95	40.17	4.73	Male	Yes	Fri	Dinner	4	0.117750	0.117750	1
90	28.97	3.00	Male	Yes	Fri	Dinner	2	0.103555	0.103555	2
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812	1
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220	2
156	48.17	5.00	Male	No	Sun	Dinner	6	0.103799	0.103799	1
182	45.35	3.50	Male	Yes	Sun	Dinner	3	0.077178	0.077178	2
197	43.11	5.00	Female	Yes	Thur	Lunch	4	0.115982	0.115982	1
142	41.19	5.00	Male	No	Thur	Lunch	5	0.121389	0.121389	2
2、RANK()

为结果集中的每一行分配一个排名，相同的值会获得相同的排名，但会跳过之后的排名,顺序为1，2，2，4，5，5，5，8......

SQL语法：

查询每天total_bill最大的两行数据

复制代码 
SELECT * FROM (
  SELECT
    t.*,
    RANK() OVER(PARTITION BY day ORDER BY total_bill DESC) AS rn
  FROM data t
)
WHERE rn < 3
ORDER BY day, rn;

对应pandas实现：

复制代码 
In :(
    data.assign(
        rnk=data.groupby(["day"])["total_bill"].rank(
            method="first", ascending=False
        )
    )
    .query("rnk < 3")
    .sort_values(["day", "rnk"])
)
Out :
total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2	rnk
95	40.17	4.73	Male	Yes	Fri	Dinner	4	0.117750	0.117750	1.0
90	28.97	3.00	Male	Yes	Fri	Dinner	2	0.103555	0.103555	2.0
170	50.81	10.00	Male	Yes	Sat	Dinner	3	0.196812	0.196812	1.0
212	48.33	9.00	Male	No	Sat	Dinner	4	0.186220	0.186220	2.0
156	48.17	5.00	Male	No	Sun	Dinner	6	0.103799	0.103799	1.0
182	45.35	3.50	Male	Yes	Sun	Dinner	3	0.077178	0.077178	2.0
197	43.11	5.00	Female	Yes	Thur	Lunch	4	0.115982	0.115982	1.0
142	41.19	5.00	Male	No	Thur	Lunch	5	0.121389	0.121389	2.0
3、SUM()

SQL语法：

复制代码 
SELECT
    t.*,
    SUM() OVER(PARTITION BY day) AS sn
  FROM data t;

In :data['sn'] = data.groupby('day')['total_bill'].cumsum()
In :data
Out :total_bill	tip	sex	smoker	day	time	size	tip_rate	tip_rate2	sn
0	16.99	1.01	Female	No	Sun	Dinner	2	0.059447	0.059447	16.99
1	10.34	1.66	Male	No	Sun	Dinner	3	0.160542	0.160542	27.33
2	21.01	3.50	Male	No	Sun	Dinner	3	0.166587	0.166587	48.34
3	23.68	3.31	Male	No	Sun	Dinner	2	0.139780	0.139780	72.02
4	24.59	3.61	Female	No	Sun	Dinner	4	0.146808	0.146808	96.61
...	...	...	...	...	...	...	...	...	...	...
239	29.03	5.92	Male	No	Sat	Dinner	3	0.203927	0.203927	1710.73
240	27.18	2.00	Female	Yes	Sat	Dinner	2	0.073584	0.073584	1737.91
241	22.67	2.00	Male	Yes	Sat	Dinner	2	0.088222	0.088222	1760.58
242	17.82	1.75	Male	No	Sat	Dinner	2	0.098204	0.098204	1778.40
243	18.78	3.00	Female	No	Thur	Dinner	2	0.159744	0.159744	1096.33
相关推荐
XiaoLeisj20 分钟前
【MyBatis】深入解析 MyBatis XML 开发：增删改查操作和方法命名规范、@Param 重命名参数、XML 返回自增主键方法
xml·java·数据库·spring boot·sql·intellij-idea·mybatis
Python之栈3 小时前
PandasAI：当数据分析遇上自然语言处理
人工智能·python·数据分析·pandas
zhuyixiangyyds11 小时前
day21和day22学习Pandas库
笔记·学习·pandas
Yan-英杰12 小时前
【百日精通JAVA | SQL篇 | 第二篇】数据库操作
服务器·数据库·sql
百代繁华一朝都-绮罗生13 小时前
检查是否存在占用内存过大的SQL
数据库·sql
橙序研工坊18 小时前
MySQL的进阶语法7（索引-B+Tree 、Hash、聚集索引 、二级索引（回表查询）、索引的使用及设计原则
数据库·sql·mysql
oh,huoyuyan1 天前
火语言RPA--Sqlite-执行SQL
sql·sqlite·rpa
云心雨禅1 天前
解决大小写、保留字与特殊字符问题！Oracle双引号在SQL中的特殊应用
数据库·sql·oracle
GoingYoo1 天前
MySQL原理：逻辑架构
数据库·sql·mysql
爱的叹息1 天前
针对 SQL 查询中 IN 子句性能优化 以及 等值 JOIN 和不等值 JOIN 对比 的详细解决方案、代码示例及表格总结
数据库·sql·性能优化
热门推荐
01我决定放弃搞 Java 了02DeepSeek各版本说明与优缺点分析03如何在WPS和Word/Excel中直接使用DeepSeek功能04RAG 实践- Ollama+RagFlow 部署本地知识库05本地化部署AI知识库：基于Ollama+DeepSeek+AnythingLLM保姆级教程06从零安装 LLaMA-Factory 微调 Qwen 大模型成功及所有的坑07如何本地部署AI智能体平台，带你手搓一个AI Agent08苍穹外卖面试总结09DeepSeek RAGFlow构建本地知识库系统10【芯片封测学习专栏 -- D2D 和 C2C 之间的区别】