数据分析(三)Pandas数据结构基础1.Series

比眉伴天荒 2023-07-10 14:09 17阅读 0赞

安装更新pandas:pip install pandas --upgrade i https://mirrors.163.com/pypi/simple/

官方[https://packaging.python.org/tutorials/installing-packages/][https_packaging.python.org_tutorials_installing-packages]

本文来自网易微专业数据分析之pandas学习笔记

### 为什么用Pandas ###

1.  **快速高效的数据结构**
2.  **智能的数据处理功能**
3.  **方便的文件存取功能**
4.  **科研及商业应用广泛**

### 两个基本的数据结构: ###

**Series和DataFrame**

![20200302222840104.png][]

### 安装 ###

pip install pandas    \#windows

### 导入pandas模块 ###

import pandas as pd

# Pandas的基础类型1---Series #

### **创建一个Series类型的数据:** ###

**1.Series()Series()里直接填一个由数字组成列表,**

data=pd.Series(\[1,3,5,7\])

**2.或设变量:**

list\_data=\[1,3,5,7\]

**data=pd.Series(list\_data)**

### **获取Series数据的值.values:** ###

**print(data.values)**

### **获取Series数据的索引值.index** ###

**print(data.index)**

>>> import pandas as pd
    >>> data=pd.Series([1,2,3,4])
    >>> print(data.values)
    [1 2 3 4]
    >>> print(data.index)
    RangeIndex(start=0, stop=4, step=1)
    >>>

**创建特殊的索引值(非数字时)**

list\_data=\[1,3,5,7\]  
data=pd.Series(list\_data,index=\['a','j','k','z'\])  
print(data)

>>> list_data=[1,3,5,7]
    >>> data=pd.Series(list_data,index=['a','j','k','z'])
    >>> print(data)
    a    1
    j    3
    k    5
    z    7
    dtype: int64

### **修改索引值:** ###

data.index=\['j','z','x','k'\]  
print(data)

>>> data.index=['j','z','x','k']
    >>> print(data)
    j    1
    z    3
    x    5
    k    7
    dtype: int64

### **获取Series数据的长度** ###

print(len(data))

### **获取数组中某个数据**   print(data\['a'\])     \#rint(data\[' j'\])   \#输入其索引号 ###

输出:1

### **获取数组中多个数据** ###

print(data\[0:2\])   \#以**切片**的方式,索引位连续的时候用  
或者:  
print(data\[\['z','k'\]\])    \#不只一个时,里面用一个**列表**,以逗号隔开,标出任意索引位,不连续时用列表,挑着用

### **计算重复元素出现的次数   \#很常用** ###

print(data.value\_counts())

![20200303085559144.png][]

### ** 判断某个索引值是否存在** ###

'a' **in** data

![20200303090222492.png][]

*与python字典相似,但却是有序的.*

### **从字典创建一个SeriesSeries类型的数据  \#很常用** ###

![2020030310424623.png][]

**给数据传入索引值(相当于给字典增加一个key)**

dict\_data=\{'Beijing': 1000, 'Shanghai': 800, 'Shenzhen': 500\}  
index\_list=\['Beijing', 'Shanghai', 'Shenzhen','Guangzhou'\]  
**data=pd.Series(dict\_data,*index*=index\_list)**  
![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70][]  
(在列表中单引号和双引号可同时用,无异常)

**检测哪些数据是缺失的(空的)**  
print(data**.isnull()**)

**检测非空**用**.notnull()**

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70 1][]

### 数组运算 ###

**乘:print(data\*2)**  
*即数组内各value均乘以2,索引是不变的*

**Series数组支持numpy数组运算**  
*把此data当成numpy的数组进行运算*  
print(np.square(data))  (求平方)

**数组相加**  
*自动把两数组中各索引值相同的进行value的相加,无匹配的依然保留.*

?如何添加数组元素?

**设置Series数组对象的名称name及其索引的名称  
data.name**='City Data'  
**data.index.name**='City'  
![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70 2][]

(本节Series完,下一节DataFrame:[https://blog.csdn.net/ahmcwt/article/details/104630629][https_blog.csdn.net_ahmcwt_article_details_104630629])

[https_packaging.python.org_tutorials_installing-packages]: https://packaging.python.org/tutorials/installing-packages/
[20200302222840104.png]: https://img-blog.csdnimg.cn/20200302222840104.png
[20200303085559144.png]: https://img-blog.csdnimg.cn/20200303085559144.png
[20200303090222492.png]: https://img-blog.csdnimg.cn/20200303090222492.png
[2020030310424623.png]: https://img-blog.csdnimg.cn/2020030310424623.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20200303121805382.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20200303123853705.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20200303130728952.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2FobWN3dA==,size_16,color_FFFFFF,t_70
[https_blog.csdn.net_ahmcwt_article_details_104630629]: https://blog.csdn.net/ahmcwt/article/details/104630629