python数据挖掘实验一：数据分析及可视化应用

1.导入pandas包

import pandas as pd

2.根据给定的原始数据集创建一个DataFrame类型对象df

data={
    'id':[10001,10002,10003,10004,10005,10006,10007,10008,10009,10010],
    'name':['LY','CE','ZS','LS','WU','ZL','SQ','ZB','WJ','ZS'],
    'sex':['F','M','M','F','F','F','M','F','M','M'],
    'age':[18,20,36,47,13,25,32,45,13,16],
    'address':['California','Texas','Florida','California','Texas','Florida','California','Florida','Texas','Texas'],
    'Mjob':['at_home','services','at_home','services','student','at_home','teacher','health','student','student'],
    'Ojob':['health','teacher','teacher','health','other','teacher','health','health','other','at_home']
}
#columns=['id','name','sex','age','address','Mjob','Ojob']，列标签
df=pd.DataFrame(data,columns=['id','name','sex','age','address','Mjob','Ojob'])

3.查看df

display(df)

4.查看df前5行和后5行

#方法一
print("前五行数据为：\n",df[:5])#左闭右开
print("后五行数据为：\n",df[5:])
#方法二
print("前五行数据为：\n",df.head())#head(n)，n不写时默认获取前5行，n为具体数据时则为获取前n行
print("后五行数据为：\n",df.tail())#tail(n)，n不写时默认获取后5行，n为具体数据时则为获取后n行

 

5.查看行和列的索引值

print("行的索引值为：",df.index)
print("列的索引值为：",df.columns)

 6.分别查看name、sex列的值

print("name列的值：\n",df['name'])
print("sex列的值：\n",df['sex'])

 7.同时查看name和sex列的值

#方法一
display(df.loc[:,['name','sex']])
#方法二
display(df.iloc[:,[1,2]])

8.连续切片，取0-5行、0-5列、左闭右开的所有数据

display(df.iloc[0:5,0:5])

9.筛选出第2-5行的第1、3、5列的所有数据

display(df.iloc[1:5,[0,2,4]])

10.分别筛选出Mjob列值不为”student”和值为“student”的所有数据

display(df[df['Mjob']=='student'])
display(df[df['Mjob']!='student'])

11.定义函数gender_count,对性别进行标识，M用1表示，F用0表示

def gender_count(x):
     if x=='M':
         return 1
     else:
         return 0
count=df['sex'].apply(gender_count)

12.df增加一列gender_n，调用函数gender_count对性别进行标识

df.insert(loc=3,column="gender_n",value=count[:])
display(df)

13.按Mjob分类，求每一种职业男性的占比

a = df.groupby(['Mjob']).gender_n.sum()/df['Mjob'].value_counts()*100
print(a)

14.将占比值从高到低的顺序排列

a2=a.sort_values(ascending=False)
print(a2)

15.求Mjob列不同值的个数

print(df.groupby('Mjob')['Mjob'].count())

16.获取每种职业对应的最大和最小的用户年龄

print("每种职业的最小用户年龄：\n",df.groupby(['Mjob','Ojob'])['age'].min())
print("每种职业的最大用户年龄：\n",df.groupby(['Mjob','Ojob'])['age'].max())

17.删除最后一列gender_n（就地修改）

df.drop(columns='gender_n',inplace=True)
display(df)

18.将数据列Mjob和Ojob中的所有数据实现首字母大写

for col in ['Mjob','Ojob']:
     df[col]=df[col].str.title()
display(df.loc[:,['Mjob','Ojob']])

19.设定id列为行索引

display(df.set_index('id'))

20.使用直方图可视化不同年龄段区间[10,20,30,40,50]的人数

import matplotlib.pyplot as plt
x = df['age'].values
plt.hist(x,bins=[10,20,30,40,50])
plt.ylim(0,5)
plt.show()

 

21.使用饼形图可视化不同类Mjob人员的占比

import matplotlib.pyplot as plt
import numpy as np
mjob=np.array(df.groupby('Mjob')['Mjob'].count().index)
num=np.array(df.groupby('Mjob')['Mjob'].count()[:])
plt.pie(num,labels=mjob,autopct='%1.0f%%')
plt.show()

22.使用柱状图可视化不同类Ojob人员的数量

import matplotlib.pyplot as plt
import numpy as np
ojob=np.array(df.groupby('Ojob')['Ojob'].count().index)
num=np.array(df.groupby('Ojob')['Ojob'].count()[:])
plt.bar(ojob,num)
plt.show()