Pandas数据分析实战：Series、DataFrame操作、数据清洗与可视化

一. 实验目的

(1) 掌握Series和DataFrame的创建；

(2) 熟悉pandas数据清洗和数据分析的常用操作；

(3) 掌握使用matplotlib库画图的基本方法。

二. 实验平台

(1) 操作系统：Windows系统；

(2) Python版本：3.8.7

三. 实验步骤

1. 基础练习

import pandas as pd
import numpy as np

# 1.创建Series
language = pd.Series(['Python', 'C', 'Scala', 'Java', 'GO', 'Scala', 'SQL', 'PHP', 'Python'])
print(language)

# 2.创建随机整型Series
score = pd.Series(np.random.randint(60, 100, len(language)))
print(score)

# 3.创建DataFrame
df = pd.DataFrame({'language': language, 'score': score})
print(df)

# 4.输出前4行数据
print(df.head(4))

# 5.输出language字段为Python的行
print(df[df['language'] == 'Python'])

# 6.按照score字段的值进行升序排序
print(df.sort_values('score'))

# 7.统计每种编程语言出现的次数
print(df['language'].value_counts())

输出结果：

0    Python
1         C
2     Scala
3      Java
4        GO
5     Scala
6       SQL
7       PHP
8    Python
dtype: object

0    87
1    75
2    89
3    68
4    84
5    88
6    86
7    99
8    92
dtype: int32

  language  score
0   Python     87
1        C     75
2    Scala     89
3     Java     68
4       GO     84
5    Scala     88
6      SQL     86
7      PHP     99
8   Python     92

  language  score
0   Python     87
1        C     75
2    Scala     89
3     Java     68

  language  score
0   Python     87
8   Python     92

  language  score
3     Java     68
1        C     75
6      SQL     86
4       GO     84
2    Scala     89
0   Python     87
8   Python     92
5    Scala     88
7      PHP     99

Scala     2
Python    2
C         1
SQL       1
GO        1
PHP       1
Java      1
Name: language, dtype: int64

2. 数据清洗练习

import pandas as pd
import numpy as np

# 1.创建DataFrame
df = pd.DataFrame({'name': ['Tom', 'Jerry', 'Mike', 'Zhangsan', 'Lisi', 'Wangwu'],
                   'age': [20, 21, 22, 23, 24, 25],
                   'gender': ['male', 'male', 'female', 'male', 'female', 'male'],
                   'score': [98, 85, 90, 75, 67, 88],
                   'city': ['Beijing', 'Shanghai', 'Guangzhou', 'Shenzhen', 'Wuhan', 'Chengdu'],
                   'nation': ['China', 'China', 'USA', 'China', 'China', 'China']})

# 2.删除名字为Wangwu的行
df = df[df['name'] != 'Wangwu']
print(df)

# 3.将city字段修改为location
df = df.rename(columns={'city': 'location'})
print(df)

# 4.统计每个nation的平均分
print(df.groupby('nation')['score'].mean())

# 5.将age字段按照以下规则映射为新的age字段：
# 20-22岁映射为young，23-25岁映射为middle
df['age'] = pd.cut(df['age'], bins=[20, 22, 25], labels=['young', 'middle'])
print(df)

输出结果：

       name  age  gender  score    city nation
0       Tom   20    male     98  Beijing  China
1     Jerry   21    male     85  Shanghai  China
2      Mike   22  female     90  Guangzhou   USA
3  Zhangsan   23    male     75  Shenzhen  China
4      Lisi   24  female     67     Wuhan  China

       name     age  gender  score  location nation
0       Tom   young    male     98   Beijing  China
1     Jerry   young    male     85  Shanghai  China
2      Mike   young  female     90  Guangzhou   USA
3  Zhangsan  middle    male     75  Shenzhen  China
4      Lisi  middle  female     67     Wuhan  China

nation
China    79.333333
USA      90.000000
Name: score, dtype: float64

       name     age  gender  score  location nation
0       Tom   young    male     98   Beijing  China
1     Jerry   young    male     85  Shanghai  China
2      Mike   young  female     90  Guangzhou   USA
3  Zhangsan  middle    male     75  Shenzhen  China
4      Lisi  middle  female     67     Wuhan  China

3. 数据分析练习

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# 1.创建DataFrame
df = pd.DataFrame({'name': ['Tom', 'Jerry', 'Mike', 'Zhangsan', 'Lisi', 'Wangwu'],
                   'age': [20, 21, 22, 23, 24, 25],
                   'gender': ['male', 'male', 'female', 'male', 'female', 'male'],
                   'score': [98, 85, 90, 75, 67, 88],
                   'city': ['Beijing', 'Shanghai', 'Guangzhou', 'Shenzhen', 'Wuhan', 'Chengdu'],
                   'nation': ['China', 'China', 'USA', 'China', 'China', 'China']})

# 2.绘制成绩分布直方图
plt.hist(df['score'], bins=range(60, 100, 5))
plt.xlabel('score')
plt.ylabel('count')
plt.title('score distribution')
plt.show()

# 3.绘制不同性别的成绩箱线图
df.boxplot(column='score', by='gender')
plt.xlabel('gender')
plt.ylabel('score')
plt.title('')
plt.suptitle('')
plt.show()

# 4.绘制不同年龄段的成绩折线图
df.groupby('age')['score'].mean().plot(kind='line')
plt.xlabel('age')
plt.ylabel('score')
plt.title('age-score relationship')
plt.show()

输出结果：