Python 数据分析：让你像写 Sql 语句一样，使用 Pandas 做数据分析
Python 数据分析：让你像写 Sql 语句一样，使用 Pandas 做数据分析
一、加载数据
import pandas as pd
import numpy as np
url = ('https://raw.github.com/pandas-dev/pandas/master/pandas/tests/data/tips.csv')
tips = pd.read_csv(url)
output = tips.head()
Output:
total_bill tip sex smoker day time size
0 16.99 1.01 Female No Sun Dinner 2
1 10.34 1.66 Male No Sun Dinner 3
2 21.01 3.50 Male No Sun Dinner 3
3 23.68 3.31 Male No Sun Dinner 2
4 24.59 3.61 Female No Sun Dinner 4
二、SELECT 的使用方式
sql 语句： SELECT total_bill, tip, smoker, time FROM tips LIMIT 5;。
output = tips[['total_bill', 'tip', 'smoker', 'time']].head(5)
Output:
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
三、WHERE 的使用方式
1. 举个栗子
sql 语句： SELECT * FROM tips WHERE time = 'Dinner' LIMIT 5;
output = tips[tips['time'] == 'Dinner'].head(5)
# 或者
output = tips.query("time == 'Dinner'").head(5)
Output:
total_bill tip sex smoker day time size
0 16.99 1.01 Female No Sun Dinner 2
1 10.34 1.66 Male No Sun Dinner 3
2 21.01 3.50 Male No Sun Dinner 3
3 23.68 3.31 Male No Sun Dinner 2
4 24.59 3.61 Female No Sun Dinner 4
2. 比较运算符：等于 ==、 大于 >、 大于等于 >=、小于等于 <=、不等于 !=
2.1 等于 ==
sql 语句：SELECT * FROM tips WHERE time = 'Dinner';。
output = tips[(tips['time'] == 'Dinner')]
2.2 大于 >
sql 语句：SELECT * FROM tips WHERE tip > 5.00;。
output = tips[(tips['tip'] > 5.00)]
2.3 大于等于 >=
sql 语句：SELECT * FROM tips WHERE tip >= 5.00;。
output = tips[(tips['size'] >= 5)]
2.4 小于等于 <=
sql 语句：SELECT * FROM tips WHERE tip <= 5.00;。
output = tips[(tips['size'] <= 5)]
2.5 不等于 !=
sql 语句：SELECT * FROM tips WHERE tip <> 5.00;。
output = tips[(tips['size'] != 5)]
3. 逻辑运算符：且 &、或 |、非 -
3.1 且 &
sql 语句：SELECT * FROM tips WHERE time = 'Dinner' AND tip > 5.00;
output = tips[(tips['time'] == 'Dinner') & (tips['tip'] > 5.00)]
3.2 或 |
sql 语句：SELECT * FROM tips WHERE size >= 5 OR total_bill > 45;。
output = tips[(tips['size'] >= 5) | (tips['total_bill'] > 45)]
3.3 非 -
sql 语句：SELECT * FROM tips WHERE not (size <> 5 AND size > 4);
output = df[-((df['size'] != 5) & (df['size'] > 4))]
4. Null 的判断
这里重新定义一个包含 NaN 数据的 DataFrame。
frame = pd.DataFrame({
'col1': ['A', 'B', np.NaN, 'C', 'D'],
'col2': ['F', np.NaN, 'G', 'H', 'I']
})
output = frame
Output:
col1 col2
0 A F
1 B NaN
2 NaN G
3 C H
4 D I
4.1 判断列是 Null
sql 语句：SELECT * FROM frame WHERE col2 IS NULL;。
output = frame[frame['col2'].isna()]
Output:
col1 col2
1 B NaN
4.2 判断列不是 Null
sql 语句：SELECT * FROM frame WHERE col1 IS NOT NULL;。
output = frame[frame['col1'].notna()]
Output:
col1 col2
0 A F
1 B NaN
3 C H
4 D I
5. In、Like 操作
5.1 In
sql 语句：SELECT * FROM tips WHERE siez in (5, 6);。
output = tips[tips['size'].isin([2, 5])]
5.2 Like
sql 语句：SELECT * FROM tips WHERE time like 'Din%';。
output = tips[tips.time.str.contains('Din*')]
四、GROUP BY 的使用方式
sql 语句：SELECT sex, count(*) FROM tips GROUP BY sex;
output = tips.groupby('sex').size()
# 获取相应的结果
output['Male']
output['Female']
output = tips.groupby('sex').count()
# 获取相应的结果
output['tip']['Female']
output = tips.groupby('sex')['total_bill'].count()
# 获取相应的结果
output['Male']
output['Female']
sql 语句：SELECT day, AVG(tip), COUNT(*) FROM tips GROUP BY day;
output = tips.groupby('day').agg({'tip': np.mean, 'day': np.size})
# 获取相应的结果
output['day']['Fri']
output['tip']['Fri']
sql 语句：SELECT smoker, day, COUNT(*), AVG(tip) FROM tips GROUP BY smoker, day;
output = tips.groupby(['smoker', 'day']).agg({'tip': [np.size, np.mean]})
# 获取相应的结果
output['tip']['size']['No']['Fri']
sql 语句：SELECT tip, count(distinct sex) FROM tips GROUP BY tip;
output = tips.groupby('tip').agg({'sex': pd.Series.nunique})
五、JOIN 连接的使用方式
定义两个 DataFrame。
df1 = pd.DataFrame({'key': ['A', 'B', 'C', 'D'], 'value': np.random.randn(4)})
df2 = pd.DataFrame({'key': ['B', 'D', 'D', 'E'], 'value': np.random.randn(4)})
1. 内连接 Inner Join
sql 语句：SELECT * FROM df1 INNER JOIN df2 ON df1.key = df2.key;
output = pd.merge(df1, df2, on='key')
# 或
indexed_df2 = df2.set_index('key')
pd.merge(df1, indexed_df2, left_on='key', right_index=True)
2. 左连接 Left Outer Join
sql 语句：SELECT * FROM df1 LEFT OUTER JOIN df2 ON df1.key = df2.key;
output = pd.merge(df1, df2, on='key', how='left')
# 或
output = df1.join(df2, on='key', how='left')
3. 右连接 Right Join
sql 语句：SELECT * FROM df1 RIGHT OUTER JOIN df2 ON df1.key = df2.key;
output = pd.merge(df1, df2, on='key', how='right')
4. 全连接 Full Join
sql 语句：SELECT * FROM df1 FULL OUTER JOIN df2 ON df1.key = df2.key;
output = pd.merge(df1, df2, on='key', how='outer')
五、UNION 的使用方式
df1 = pd.DataFrame({'city': ['Chicago', 'San Francisco', 'New York City'], 'rank': range(1, 4)})
df2 = pd.DataFrame({'city': ['Chicago', 'Boston', 'Los Angeles'], 'rank': [1, 4, 5]})
sql 语句：SELECT city, rank FROM df1 UNION ALL SELECT city, rank FROM df2;
output = pd.concat([df1, df2])
sql 语句：SELECT city, rank FROM df1 UNION SELECT city, rank FROM df2;
output = pd.concat([df1, df2]).drop_duplicates()
六、与 SQL 等价的其他语法
1. 去重 Distinct
sql 语句：SELECT DISTINCT sex FROM tips;
output = tips.drop_duplicates(subset=['sex'], keep='first', inplace=False)
2. 修改列别名 As
sql 语句：SELECT total_bill AS total, sex AS xes FROM tips;
output = tips.rename(columns={'total_bill': 'total', 'sex': 'xes'}, inplace=False)
3. Limit 与 Offset
sql 语句：SELECT * FROM tips ORDER BY tip DESC LIMIT 10 OFFSET 5;
output = tips.nlargest(10 + 5, columns='tip').tail(10)
4. 每个 Group 的前几行
sql 语句：
SELECT * FROM (
SELECT
t.*,
ROW_NUMBER() OVER(PARTITION BY day ORDER BY total_bill DESC) AS rn
FROM tips t
)
WHERE rn < 3
ORDER BY day, rn;
output = tips.assign(rn=tips.sort_values(['total_bill'], ascending=False).\
groupby(['day']).cumcount() + 1).\
query('rn < 3').\
sort_values(['day', 'rn'])
七、Update 的使用方式
sql 语句：UPDATE tips SET tip = tip*2 WHERE tip < 2;
output = tips.loc[tips['tip'] < 2, 'tip'] *= 2
八、Delete 的使用方式
sql 语句：DELETE FROM tips WHERE tip > 9;
output = tips = tips.loc[tips['tip'] <= 9]
