数据分析厉器---pandas入门

不念不忘少年蓝@ 2022-12-20 06:08 200阅读 0赞

*  **pandas**是python的非常好用的数据分析的库，
 *  pandas的优势  
    增强图表可读性  
    便捷的数据处理能力  
    读取文件方便  
    封装了Matplotlib、Numpy的画图和计算
 *  ![pandas优势][pandas]

# series #

## series的创建 ##

*  Series是一个类似于一维数组的数据结构，它能够保存任何类型的数据，比如整数、字符串、浮点数等，主要由一组数据和与之相关的索引两部分构成。

import pandas as pd
    import numpy as np

*  参数：
    
     *  data：传入的数据，可以是ndarray、list等
     *  index：索引，必须是唯一的，且与数据的长度相等。如果没有传入索引参数，则默认会自动创建一个从0-N的整数索引。
     *  dtype：数据的类型

pd.Series(np.arange(9))

0    0
    1    1
    2    2
    3    3
    4    4
    5    5
    6    6
    7    7
    8    8
    dtype: int32

pd.Series([1.2, 2.3, 4.5, 5.0], index=[1, 2, 3, 4])

1    1.2
    2    2.3
    3    4.5
    4    5.0
    dtype: float64

color_count = pd.Series({
        "red":10, "green":20, "blue":100})

color_count

red       10
    green     20
    blue     100
    dtype: int64

## series的属性 ##

*  为了更方便地操作Series对象中的索引和数据，Series中提供了两个属性index和values

color_count.index

Index(['red', 'green', 'blue'], dtype='object')

color_count.values

array([ 10,  20, 100], dtype=int64)

color_count[0]

color_count[1]

# DataFrame #

## DataFrame创建 ##

*  DataFrame是一个类似于二维数组或表格(如excel)的对象，既有行索引，又有列索引
    
     *  行索引，表明不同行，横向索引，叫index，0轴，axis=0
     *  列索引，表名不同列，纵向索引，叫columns，1轴，axis=1

pd.DataFrame(np.random.randn(2,3))

score = np.random.randint(40, 100, (10, 5))

score

array([[47, 80, 91, 99, 80],
           [48, 79, 94, 94, 50],
           [76, 66, 88, 81, 64],
           [62, 61, 90, 97, 83],
           [49, 98, 81, 69, 52],
           [77, 70, 75, 97, 56],
           [97, 67, 45, 40, 49],
           [78, 99, 64, 69, 98],
           [93, 58, 50, 46, 99],
           [49, 70, 64, 68, 72]])

score_df = pd.DataFrame(score)

score_df

subjects = ["语文", "数学", "英语", "政治", "体育"]
    
    stu = ["同学"+ str(i) for i in range(score_df.shape[0])]
    
    data = pd.DataFrame(score, columns=subjects, index=stu)

stu

['同学0', '同学1', '同学2', '同学3', '同学4', '同学5', '同学6', '同学7', '同学8', '同学9']

data

## DataFrame的属性 ##

data.shape

(10, 5)

data.index

Index(['同学0', '同学1', '同学2', '同学3', '同学4', '同学5', '同学6', '同学7', '同学8', '同学9'], dtype='object')

data.columns

Index(['语文', '数学', '英语', '政治', '体育'], dtype='object')

data.values

data.T

data

data.head()

data.head(3)

data.tail(3)

## DataFrame索引值的设置 ##

stu = ["同学_"+ str(i) for i in range(score_df.shape[0])]
    
    data.index = stu

# stu

data

# data.index[2] = "同学__"

data.reset_index()

<table> 
 <thead> 
  <tr> 
   <th></th> 
   <th>index</th> 
   <th>语文</th> 
   <th>数学</th> 
   <th>英语</th> 
   <th>政治</th> 
   <th>体育</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <th>0</th> 
   <td>同学_0</td> 
   <td>47</td> 
   <td>80</td> 
   <td>91</td> 
   <td>99</td> 
   <td>80</td> 
  </tr> 
  <tr> 
   <th>1</th> 
   <td>同学_1</td> 
   <td>48</td> 
   <td>79</td> 
   <td>94</td> 
   <td>94</td> 
   <td>50</td> 
  </tr> 
  <tr> 
   <th>2</th> 
   <td>同学_2</td> 
   <td>76</td> 
   <td>66</td> 
   <td>88</td> 
   <td>81</td> 
   <td>64</td> 
  </tr> 
  <tr> 
   <th>3</th> 
   <td>同学_3</td> 
   <td>62</td> 
   <td>61</td> 
   <td>90</td> 
   <td>97</td> 
   <td>83</td> 
  </tr> 
  <tr> 
   <th>4</th> 
   <td>同学_4</td> 
   <td>49</td> 
   <td>98</td> 
   <td>81</td> 
   <td>69</td> 
   <td>52</td> 
  </tr> 
  <tr> 
   <th>5</th> 
   <td>同学_5</td> 
   <td>77</td> 
   <td>70</td> 
   <td>75</td> 
   <td>97</td> 
   <td>56</td> 
  </tr> 
  <tr> 
   <th>6</th> 
   <td>同学_6</td> 
   <td>97</td> 
   <td>67</td> 
   <td>45</td> 
   <td>40</td> 
   <td>49</td> 
  </tr> 
  <tr> 
   <th>7</th> 
   <td>同学_7</td> 
   <td>78</td> 
   <td>99</td> 
   <td>64</td> 
   <td>69</td> 
   <td>98</td> 
  </tr> 
  <tr> 
   <th>8</th> 
   <td>同学_8</td> 
   <td>93</td> 
   <td>58</td> 
   <td>50</td> 
   <td>46</td> 
   <td>99</td> 
  </tr> 
  <tr> 
   <th>9</th> 
   <td>同学_9</td> 
   <td>49</td> 
   <td>70</td> 
   <td>64</td> 
   <td>68</td> 
   <td>72</td> 
  </tr> 
 </tbody> 
</table>

data.reset_index(drop=True)

df = pd.DataFrame({
        'month': [1, 4, 7, 10],
                        'year': [2012, 2014, 2013, 2014],
                        'sale':[55, 40, 84, 31]})

<table> 
 <thead> 
  <tr> 
   <th></th> 
   <th>month</th> 
   <th>year</th> 
   <th>sale</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <th>0</th> 
   <td>1</td> 
   <td>2012</td> 
   <td>55</td> 
  </tr> 
  <tr> 
   <th>1</th> 
   <td>4</td> 
   <td>2014</td> 
   <td>40</td> 
  </tr> 
  <tr> 
   <th>2</th> 
   <td>7</td> 
   <td>2013</td> 
   <td>84</td> 
  </tr> 
  <tr> 
   <th>3</th> 
   <td>10</td> 
   <td>2014</td> 
   <td>31</td> 
  </tr> 
 </tbody> 
</table>

df.set_index("year")

<table> 
 <thead> 
  <tr> 
   <th></th> 
   <th>month</th> 
   <th>sale</th> 
  </tr> 
  <tr> 
   <th>year</th> 
   <th></th> 
   <th></th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <th>2012</th> 
   <td>1</td> 
   <td>55</td> 
  </tr> 
  <tr> 
   <th>2014</th> 
   <td>4</td> 
   <td>40</td> 
  </tr> 
  <tr> 
   <th>2013</th> 
   <td>7</td> 
   <td>84</td> 
  </tr> 
  <tr> 
   <th>2014</th> 
   <td>10</td> 
   <td>31</td> 
  </tr> 
 </tbody> 
</table>

df = df.set_index(["year", "month"])

<table> 
 <thead> 
  <tr> 
   <th></th> 
   <th></th> 
   <th>sale</th> 
  </tr> 
  <tr> 
   <th>year</th> 
   <th>month</th> 
   <th></th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <th>2012</th> 
   <th>1</th> 
   <td>55</td> 
  </tr> 
  <tr> 
   <th>2014</th> 
   <th>4</th> 
   <td>40</td> 
  </tr> 
  <tr> 
   <th>2013</th> 
   <th>7</th> 
   <td>84</td> 
  </tr> 
  <tr> 
   <th>2014</th> 
   <th>10</th> 
   <td>31</td> 
  </tr> 
 </tbody> 
</table>

# MultiIndex与Panel #

## MultiIndex ##

df.index

MultiIndex([(2012,  1),
                (2014,  4),
                (2013,  7),
                (2014, 10)],
               names=['year', 'month'])

df.index.names

FrozenList(['year', 'month'])

df.index.levels

FrozenList([[2012, 2013, 2014], [1, 4, 7, 10]])

arrays = [[1, 1, 2, 2], ["r", "b", "r","b"]]
    pd.MultiIndex.from_arrays(arrays, names=("num", "col"))

MultiIndex([(1, 'r'),
                (1, 'b'),
                (2, 'r'),
                (2, 'b')],
               names=['num', 'col'])

## panel ##

p = pd.Panel(data=np.arange(24).reshape(4,3,2),
                     items=list('ABCD'),
                     major_axis=pd.date_range('20130101', periods=3),
                     minor_axis=['first', 'second'])

c:\python\lib\site-packages\ipykernel_launcher.py:1: FutureWarning: The Panel class is removed from pandas. Accessing it from the top-level namespace will also be removed in the next version
      """Entry point for launching an IPython kernel.
    
    
    
    ---------------------------------------------------------------------------
    
    TypeError                                 Traceback (most recent call last)
    
    <ipython-input-43-78e9fbdccf04> in <module>
          2                  items=list('ABCD'),
          3                  major_axis=pd.date_range('20130101', periods=3),
    ----> 4                  minor_axis=['first', 'second'])
    
    
    TypeError: Panel() takes no arguments

p[:, :, "second"]

p[:, :, "first"]

p["A", :, :]

[pandas]: /images/20221120/73d287411b30478a8fabd5cf4bc5b9d8.png