Pandas入门

1 Pandas介绍

• 专门用于数据挖掘的开源python库
• 以Numpy为基础,借力Numpy模块在计算方面性能优势
• 基于matplotlib,能够简便的画图
• 独特的数据结构

Pandas 是一种数据处理工具，它是以下面三组词汇组成

panel + data + analysis

panel：面板数据是计量经济学中常用，一般表示三维数据

2 为什么使用Pandas

• 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

3 DataFrame

从图中可以看到DataFrame由三部分组成

• column label即列标签
• index label即行标签
• data即数据

那么标记轴axis表示什么呢，axis有两种取值，分别是1和0，其中1表示横轴，方向从左到右；0表示纵轴，方向从上到下。

import numpy as np
#创建一个符合正态分布的10个股票5天的涨跌幅数据
stock_change =np.random.normal(0,1,(10,5))
stock_change

array([[ 0.55780125,  0.47366431,  0.58266456, -0.29946146, -0.03390217],
       [-0.24385523,  0.08817049,  1.38707642, -0.57688673, -0.34760394],
       [ 0.95549368,  0.9414475 , -1.25056314,  0.18178455, -0.29557978],
       [ 1.61507705,  1.99202826,  2.80758189,  0.03192688, -0.57838353],
       [ 1.4956878 , -1.23262134,  2.50024192, -0.58850329,  0.7102027 ],
       [-0.662319  ,  1.76285879,  1.51286286, -0.53192944, -0.47949495],
       [ 0.73735599,  0.48964047, -1.32854508, -0.07826431,  0.36766669],
       [ 0.79199457, -1.74662017, -0.334844  , -1.47935611, -0.12609656],
       [-1.77942406, -1.67284383, -0.90279781,  0.06015451,  0.66952752],
       [-1.93908274, -1.93232172,  0.8559445 ,  1.13113002,  1.33307564]])

import pandas as pd
pd.DataFrame(stock_change)

	0	1	2	3	4
0	0.557801	0.473664	0.582665	-0.299461	-0.033902
1	-0.243855	0.088170	1.387076	-0.576887	-0.347604
2	0.955494	0.941448	-1.250563	0.181785	-0.295580
3	1.615077	1.992028	2.807582	0.031927	-0.578384
4	1.495688	-1.232621	2.500242	-0.588503	0.710203
5	-0.662319	1.762859	1.512863	-0.531929	-0.479495
6	0.737356	0.489640	-1.328545	-0.078264	0.367667
7	0.791995	-1.746620	-0.334844	-1.479356	-0.126097
8	-1.779424	-1.672844	-0.902798	0.060155	0.669528
9	-1.939083	-1.932322	0.855944	1.131130	1.333076

给股票数据增加行列索引

pd.DataFrame(stock_change,行索引，列索引)

#  添加行索引 列表解析
stock = ["股票{}".format(i) for i in range(10)]

pd.DataFrame(stock_change,stock)

	0	1	2	3	4
股票0	0.557801	0.473664	0.582665	-0.299461	-0.033902
股票1	-0.243855	0.088170	1.387076	-0.576887	-0.347604
股票2	0.955494	0.941448	-1.250563	0.181785	-0.295580
股票3	1.615077	1.992028	2.807582	0.031927	-0.578384
股票4	1.495688	-1.232621	2.500242	-0.588503	0.710203
股票5	-0.662319	1.762859	1.512863	-0.531929	-0.479495
股票6	0.737356	0.489640	-1.328545	-0.078264	0.367667
股票7	0.791995	-1.746620	-0.334844	-1.479356	-0.126097
股票8	-1.779424	-1.672844	-0.902798	0.060155	0.669528
股票9	-1.939083	-1.932322	0.855944	1.131130	1.333076

#添加列索引(先行后列)
date = pd.date_range(start ="20180101",periods=5,freq="B")
data =pd.DataFrame(stock_change,stock,date)

结构：既有行索引，又有列索引的二维数组

	2018-01-01 00:00:00	2018-01-02 00:00:00	2018-01-03 00:00:00	2018-01-04 00:00:00	2018-01-05 00:00:00
股票0	0.557801	0.473664	0.582665	-0.299461	-0.033902
股票1	-0.243855	0.088170	1.387076	-0.576887	-0.347604
股票2	0.955494	0.941448	-1.250563	0.181785	-0.295580
股票3	1.615077	1.992028	2.807582	0.031927	-0.578384
股票4	1.495688	-1.232621	2.500242	-0.588503	0.710203
股票5	-0.662319	1.762859	1.512863	-0.531929	-0.479495
股票6	0.737356	0.489640	-1.328545	-0.078264	0.367667
股票7	0.791995	-1.746620	-0.334844	-1.479356	-0.126097
股票8	-1.779424	-1.672844	-0.902798	0.060155	0.669528
股票9	-1.939083	-1.932322	0.855944	1.131130	1.333076

3.1 属性

• shape 形状
• index 行索引
• columns 列索引
• values NDARRAY数据
• T 转置

data.shape

(10, 5)

data.index

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

data.values

array([[ 0.55780125,  0.47366431,  0.58266456, -0.29946146, -0.03390217],
       [-0.24385523,  0.08817049,  1.38707642, -0.57688673, -0.34760394],
       [ 0.95549368,  0.9414475 , -1.25056314,  0.18178455, -0.29557978],
       [ 1.61507705,  1.99202826,  2.80758189,  0.03192688, -0.57838353],
       [ 1.4956878 , -1.23262134,  2.50024192, -0.58850329,  0.7102027 ],
       [-0.662319  ,  1.76285879,  1.51286286, -0.53192944, -0.47949495],
       [ 0.73735599,  0.48964047, -1.32854508, -0.07826431,  0.36766669],
       [ 0.79199457, -1.74662017, -0.334844  , -1.47935611, -0.12609656],
       [-1.77942406, -1.67284383, -0.90279781,  0