Pandas-数据结构-Series（二）：Series的索引【下标索引、标签索引、切片索引、布尔型索引】

青旅半醒 2023-09-29 09:02 13阅读 0赞

## 一、下标索引 ##

位置下标，类似序列

*  位置下标从0开始
 *  输出结果为numpy.float格式，
 *  可以通过float()函数转换为python float格式
 *  numpy.float与float占用字节不同

import numpy as np
    import pandas as pd
    
    s = pd.Series(np.random.rand(5))
    print('s = \n', s)
    print('\ns[0] = {0}, type(s[0]) = {1}, s[0].dtype = {2}'.format(s[0], type(s[0]), s[0].dtype))
    print('\nfloat(s[0]) = {0}, type(float(s[0]) = {1})'.format(float(s[0]), type(float(s[0]))))

打印结果：

s = 
    0    0.221458
    1    0.430786
    2    0.096305
    3    0.262580
    4    0.993637
    dtype: float64
    
    s[0] = 0.22145841095529073, type(s[0]) = <class 'numpy.float64'>, s[0].dtype = float64
    
    float(s[0]) = 0.22145841095529073, type(float(s[0]) = <class 'float'>)
    
    Process finished with exit code 0

## 二、标签索引 ##

方法类似下标索引，用\[\]表示，内写上index，注意index是字符串

*  如果需要选择多个标签的值，用\[\[\]\]来表示（相当于\[\]中包含一个列表）
 *  多标签索引结果是新的数组

import numpy as np
    import pandas as pd
    
    # 方法类似下标索引，用[]表示，内写上index，注意index是字符串
    s = pd.Series(np.random.rand(5), index=['a', 'b', 'c', 'd', 'e'])
    print('\ns = \n{0}, \n\ntype(s) = {1}'.format(s, type(s)))
    
    print("\ns['a'] = {0}, type(s['a']) = {1}, s['a'].dtype = {2}".format(s['a'], type(s['a']), s['a'].dtype))
    
    print('-' * 100)
    
    sci = s[['a', 'b', 'e']]
    print('\nsci = \n{0}, \n\ntype(sci) = {1}'.format(sci, type(sci)))

打印结果：

s = 
    a    0.648198
    b    0.356047
    c    0.375873
    d    0.811611
    e    0.070581
    dtype: float64, 
    
    type(s) = <class 'pandas.core.series.Series'>
    
    s['a'] = 0.6481978904955534, type(s['a']) = <class 'numpy.float64'>, s['a'].dtype = float64
    ----------------------------------------------------------------------------------------------------
    
    sci = 
    a    0.648198
    b    0.356047
    e    0.070581
    dtype: float64, 
    
    type(sci) = <class 'pandas.core.series.Series'>
    
    Process finished with exit code 0

## 三、切片索引 ##

注意：

*  用index做切片是末端包含
 *  下标索引做切片，和list写法一样

import numpy as np
    import pandas as pd
    
    # 注意：用index做切片是末端包含
    
    s1 = pd.Series(np.random.rand(5))
    s2 = pd.Series(np.random.rand(5), index = ['a','b','c','d','e'])
    print('s1 = \n', s1)
    print('-' * 30)
    print('s1[1:4] = \n', s1[1:4])
    print('-' * 30)
    print('s1[4] = \n', s1[4])
    print('-' * 100)
    print('s2 = \n', s2)
    print('-' * 30)
    print("s2['a':'c'] = \n", s2['a':'c'])
    print('-' * 30)
    print("s2['c'] = \n", s2['c'])
    print('-' * 30)
    print("s2[0:3] = \n", s2[0:3])
    print('-' * 30)
    print("s2[3] = \n", s2[3])
    print('-' * 30)
    print("s2[:-1] = \n", s2[:-1])
    print('-' * 30)
    print("s2[::2] = \n", s2[::2])
    print('-' * 100)

打印结果：

s1 = 
    0    0.786107
    1    0.223454
    2    0.591863
    3    0.911600
    4    0.672031
    dtype: float64
    ------------------------------
    s1[1:4] = 
    1    0.223454
    2    0.591863
    3    0.911600
    dtype: float64
    ------------------------------
    s1[4] = 
     0.6720314874669562
    ----------------------------------------------------------------------------------------------------
    s2 = 
    a    0.398821
    b    0.856505
    c    0.795255
    d    0.985476
    e    0.724451
    dtype: float64
    ------------------------------
    s2['a':'c'] = 
    a    0.398821
    b    0.856505
    c    0.795255
    dtype: float64
    ------------------------------
    s2['c'] = 
     0.7952552440685834
    ------------------------------
    s2[0:3] = 
    a    0.398821
    b    0.856505
    c    0.795255
    dtype: float64
    ------------------------------
    s2[3] = 
     0.9854755359882719
    ------------------------------
    s2[:-1] = 
    a    0.398821
    b    0.856505
    c    0.795255
    d    0.985476
    dtype: float64
    ------------------------------
    s2[::2] = 
    a    0.398821
    c    0.795255
    e    0.724451
    dtype: float64
    ----------------------------------------------------------------------------------------------------
    
    Process finished with exit code 0

## 四、布尔型索引 ##

数组做判断之后，返回的是一个由布尔值组成的新的数组

*  .isnull() / .notnull() 判断是否为空值 (None代表空值，NaN代表有问题的数值，两个都会识别为空值)
 *  布尔型索引方法：用\[判断条件\]表示，其中判断条件可以是 一个语句，或者是 一个布尔型数组！

import numpy as np
    import pandas as pd
    
    s = pd.Series(np.random.rand(3)*100)
    s[4] = None  # 添加一个空值
    print("s = \n", s)
    print('-' * 100)
    # 数组做判断之后，返回的是一个由布尔值组成的新的数组
    bs1 = s > 50
    print("bs1 = \n{0}, \ntype(bs1) = {1}, \nbs1.dtype = {2}".format(bs1, type(bs1), bs1.dtype))
    print('-' * 30)
    # .isnull() / .notnull() 判断是否为空值 (None代表空值，NaN代表有问题的数值，两个都会识别为空值)
    bs2 = s.isnull()
    bs3 = s.notnull()
    print("bs2 = \n{0}, \ntype(bs2) = {1}, bs2.dtype = {2}".format(bs2, type(bs2), bs2.dtype))
    print('-' * 30)
    print("bs3 = \n{0}, \ntype(bs3) = {1}, bs3.dtype = {2}".format(bs3, type(bs3), bs3.dtype))
    print('-' * 100)
    
    print("s[s > 50] = \n", s[s > 50])
    print('-' * 30)
    print("s[bs3] \n", s[bs3])

打印结果：

s = 
    0    11.735721
    1    62.482804
    2    51.381651
    4         None
    dtype: object
    ----------------------------------------------------------------------------------------------------
    bs1 = 
    0    False
    1     True
    2     True
    4    False
    dtype: bool, 
    type(bs1) = <class 'pandas.core.series.Series'>, 
    bs1.dtype = bool
    ------------------------------
    bs2 = 
    0    False
    1    False
    2    False
    4     True
    dtype: bool, 
    type(bs2) = <class 'pandas.core.series.Series'>, bs2.dtype = bool
    ------------------------------
    bs3 = 
    0     True
    1     True
    2     True
    4    False
    dtype: bool, 
    type(bs3) = <class 'pandas.core.series.Series'>, bs3.dtype = bool
    ----------------------------------------------------------------------------------------------------
    s[s > 50] = 
    1    62.482804
    2    51.381651
    dtype: object
    ------------------------------
    s[bs3] 
    0    11.735721
    1    62.482804
    2    51.381651
    dtype: object
    
    Process finished with exit code 0