数据挖掘工具pandas（二）DataFrame的索引

一，设置DataFrame索引值 以及 时间索引如何构造

1，设置DataFrame索引值

import numpy as np
import pandas as pd

day_data = np.random.normal(0,1,(500,507))

# 将数据变成dataframe格式
day_data1 = pd.DataFrame(day_data)

# 构造行索引列表
stock_list = ["股票"+ str(i) for i in range(day_data.shape[0])]

# 构造列索引列表
date = pd.date_range("2018-01-01",periods=day_data.shape[1],freq='B')

# 设置行、列索引
day_data2 = pd.DataFrame(day_data,index=stock_list,columns=date)
print(day_data2)

![在这里插入图片描述](https://img-blog.csdnimg.cn/20200522172859889.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1RGQVRT,size_16,color_FFFFFF,t_70)

2，时间索引如何构造

pd.date_range()：用于生成一组连续的时间序列
date_range(start=None, end=None, periods=None, freq=‘D’)

• start:开始时间
• end:结束时间
• periods:时间天数
• freq:递进单位，默认1天，频率

二，dataframe修改索引

1，直接修改

import numpy as np
import pandas as pd

day_data = np.random.normal(0,1,(500,507))
day_data1 = pd.DataFrame(day_data)
stock_list = ["股票"+ str(i) for i in range(day_data.shape[0])]
date = pd.date_range("2018-01-01",periods=day_data.shape[1],freq='B')
day_data2 = pd.DataFrame(day_data,index=stock_list,columns=date)
print(day_data2)

# 修改索引必须整体全部修改
date = ["第"+ str(i)+"天" for i in range(day_data.shape[1])]
day_data2.columns = date
print(day_data2)

2，重设索引（ reset_index() ）

import numpy as np
import pandas as pd

day_data = np.random.normal(0,1,(500,507))
day_data1 = pd.DataFrame(day_data)
stock_list = ["股票"+ str(i) for i in range(day_data.shape[0])]
date = ["第"+ str(i)+"天" for i in range(day_data.shape[1])]
day_data1 = pd.DataFrame(day_data,index=stock_list,columns=date)

# 如果reset_index设置drop为默认的False，则将行索引加入到数据中，替换的是默认行索引
# day_data1 = day_data1.reset_index()

# 如果reset_index设置drop为True，则将行索引删除，替换的是默认行索引
day_data1 = day_data1.reset_index(drop=True)

3，以某列值设为新的索引（ set_index() ）

import pandas as pd

d1 = {
    "name" : ["xiaoming","xiaofang","xiaoli"],
    "age" : [20,32,24],
    "tel" : [10086,10000,10010],
    "agent":["male","female","male"],
}
a = pd.DataFrame(d1)
print(a )
a =a.set_index(["name","age"],drop=True)
print(a)

#--------output-----------------
       name  age    tel   agent
0  xiaoming   20  10086    male
1  xiaofang   32  10000  female
2    xiaoli   24  10010    male
                tel   agent
name     age               
xiaoming 20   10086    male
xiaofang 32   10000  female
xiaoli   24   10010    male

三，dataframe按索引排序

排序有两种，一种对索引进行排序，一种对内容进行排序

• 使用df.sort_values给内容排序（默认是从小到大）
  • 单个键进行排序
  • 多个键进行排序
• 使用df.sort_index给索引进行排序

1，对内容进行排序-df.sort_values()

# encoding=utf-8
import numpy as np
import pandas as pd

day_data = np.random.normal(0,1,(500,507))
day_data1 = pd.DataFrame(day_data)
stock_list = [ int(i) for i in range(day_data.shape[0])]
date = ["第"+ str(i)+"天" for i in range(day_data.shape[1])]
day_data1 = pd.DataFrame(day_data,index=stock_list,columns=date)

# 按照内容排序，默认为True-从小到大
# 单个键进行排序
# print(day_data1.sort_values(by="第1天",ascending=True))
# 多个键进行排序
print(day_data1.sort_values(by=["第1天","第2天"],ascending=False))

2，对索引进行排序-df.sort_index()

# encoding=utf-8
import numpy as np
import pandas as pd

day_data = np.random.normal(0,1,(500,507))
day_data1 = pd.DataFrame(day_data)
stock_list = [ int(i) for i in range(day_data.shape[0])]
date = ["第"+ str(i)+"天" for i in range(day_data.shape[1])]
day_data1 = pd.DataFrame(day_data,index=stock_list,columns=date)

# 按照索引排序，默认为True-从小到大
print(day_data1.sort_index(ascending=False))