zoukankan html css js c++ java

pandas模块

pandas更多的是excel/csv文件处理,excel文件, 对numpy+xlrd模块做了一层封装

pandas的数据类型

import pandas as pd
import numpy as np

serise(现在一般不使用(一维))

df = pd.Series(np.array([1,2,3,4]))
print(df)

DataFrame多维

dates = pd.date_range('20190101', periods=6, freq='M')
print(dates)

values = np.random.rand(6, 4) * 10
print(values)

columns = ['c1','c2','c3','c3']

df = pd.DataFrame(values,index=dates,columns=columns)
print(df)

DatetimeIndex(['2019-01-31', '2019-02-28', '2019-03-31', '2019-04-30',
'2019-05-31', '2019-06-30'],
dtype='datetime64[ns]', freq='M')
[[1.16335011 4.48613539 7.68543002 2.28527564]
[8.93474708 7.31073142 8.61253719 2.50506357]
[4.88797902 6.81381968 9.0847644 8.34332396]
[6.74341716 9.32192571 9.01122189 2.93191827]
[3.83096571 3.27206377 7.25800888 1.30570883]
[2.87592228 0.17123983 4.97889883 5.4085225 ]]
c1 c2 c3 c3
2019-01-31 1.163350 4.486135 7.685430 2.285276
2019-02-28 8.934747 7.310731 8.612537 2.505064
2019-03-31 4.887979 6.813820 9.084764 8.343324
2019-04-30 6.743417 9.321926 9.011222 2.931918
2019-05-31 3.830966 3.272064 7.258009 1.305709
2019-06-30 2.875922 0.171240 4.978899 5.408523

DataFrame属性

dtype 查看数据类型
index 查看行序列或者索引
columns 查看各列的标签
values 查看数据框内的数据，也即不含表头索引的数据
describe 查看数据每一列的极值，均值，中位数，只可用于数值型数据
transpose 转置，也可用Ｔ来操作
sort_index 排序，可按行或列index排序输出
sort_values 按数据值来排序

print(df.dtypes)
print(df.index)
print(df.columns)
print(df.describe())
print(df.T)

import pandas as pd
import numpy as np
dates = pd.date_range('20190101', periods=6, freq='M')
print(dates)
values = np.random.rand(6, 4) * 10
print(values)
columns = ['c4','c2','c3','c1']
df = pd.DataFrame(values,index=dates,columns=columns)
print(df)

	c4	c2	c3	c1
2019-01-31	5.820943	8.551214	4.164049	1.268047
2019-02-28	6.809855	3.161353	1.934861	3.639872
2019-03-31	0.679617	6.166411	3.264278	3.919507
2019-04-30	2.634395	8.825472	2.345733	0.301147
2019-05-31	9.859531	9.294794	4.025121	3.545862
2019-06-30	5.566927	0.043362	5.301493	0.214879

df.T

	2019-01-31	2019-02-28	2019-03-31	2019-04-30	2019-05-31	2019-06-30
c4	5.820943	6.809855	0.679617	2.634395	9.859531	5.566927
c2	8.551214	3.161353	6.166411	8.825472	9.294794	0.043362
c3	4.164049	1.934861	3.264278	2.345733	4.025121	5.301493
c1	1.268047	3.639872	3.919507	0.301147	3.545862	0.214879

df = df.sort_index(axis=1)  # 0列,1是行
df

	c1	c2	c3	c4
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-02-28	3.639872	3.161353	1.934861	6.809855
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-04-30	0.301147	8.825472	2.345733	2.634395
2019-05-31	3.545862	9.294794	4.025121	9.859531
2019-06-30	0.214879	0.043362	5.301493	5.566927

df.sort_values('c3')

	c1	c2	c3	c4
2019-02-28	3.639872	3.161353	1.934861	6.809855
2019-04-30	0.301147	8.825472	2.345733	2.634395
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-05-31	3.545862	9.294794	4.025121	9.859531
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-06-30	0.214879	0.043362	5.301493	5.566927

取值

df['c1]

2019-01-31    1.268047
2019-02-28    3.639872
2019-03-31    3.919507
2019-04-30    0.301147
2019-05-31    3.545862
2019-06-30    0.214879
Freq: M, Name: c1, dtype: float64

df[['c1','c3']]

	c1	c3
2019-01-31	1.268047	4.164049
2019-02-28	3.639872	1.934861
2019-03-31	3.919507	3.264278
2019-04-30	0.301147	2.345733
2019-05-31	3.545862	4.025121
2019-06-30	0.214879	5.301493

df.loc['2019-01-31':'2019-02-28']

	c1	c2	c3	c4
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-02-28	3.639872	3.161353	1.934861	6.809855

df.values[1,1]

3.1613533123062734

	c1	c2	c3	c4
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-02-28	3.639872	3.161353	1.934861	6.809855
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-04-30	0.301147	8.825472	2.345733	2.634395
2019-05-31	3.545862	9.294794	4.025121	9.859531
2019-06-30	0.214879	0.043362	5.301493	5.566927

df.iloc[:,:]

	c1	c2	c3	c4
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-02-28	3.639872	3.161353	1.934861	6.809855
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-04-30	0.301147	8.825472	2.345733	2.634395
2019-05-31	3.545862	9.294794	4.025121	9.859531
2019-06-30	0.214879	0.043362	5.301493	5.566927

df[df['c1']>3]

	c1	c2	c3	c4
2019-02-28	3.639872	3.161353	1.934861	6.809855
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-05-31	3.545862	9.294794	4.025121	9.859531

值替换

df.iloc[1,1] = 1
df

	c1	c2	c3	c4
2019-01-31	1.268047	8.551214	4.164049	5.820943
2019-02-28	3.639872	1.000000	1.934861	6.809855
2019-03-31	3.919507	6.166411	3.264278	0.679617
2019-04-30	0.301147	8.825472	2.345733	2.634395
2019-05-31	3.545862	9.294794	4.025121	9.859531
2019-06-30	0.214879	0.043362	5.301493	5.566927

pandas操作表格

from io import StringIO
test_data = '''
5.1,,1.4,0.2
4.9,3.0,1.4,0.2
4.7,3.2,,0.2
7.0,3.2,4.7,1.4
6.4,3.2,4.5,1.5
6.9,3.1,4.9,
,,,
'''
print(test_data)
test_data = StringIO(test_data)  # 把test_data读入内存,相当于变成文件
print(test_data)

# 把数据读入内存,变成csv文件

5.1,,1.4,0.2
4.9,3.0,1.4,0.2
4.7,3.2,,0.2
7.0,3.2,4.7,1.4
6.4,3.2,4.5,1.5
6.9,3.1,4.9,
,,,

<_io.StringIO object at 0x000001665DD6A828>

df = pd.read_csv('test.csv', header=None) #读取文件  # header没有columns
df.columns =['c1','c2','c3','c4']
df

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

缺失值处理

df = df.dropna(axis=0) # 1列,0行
df

	c1	c2	c3	c4
1	4.9	3.0	1.4	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5

df = df.dropna(thresh=3) # 必须得有4个值
df

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN

合并处理

df1 = pd.DataFrame(np.zeros((2,3)))
df1

0	1	2
0	0.0	0.0	0.0
1	0.0	0.0	0.0

df2 = pd.DataFrame(np.ones((2,3)))
df2

	0	1	2
0	1.0	1.0	1.0
1	1.0	1.0	1.0

pd.concat((df1,df2),axis=1)  # 默认按列0,1行

	0	1	2	0	1	2
0	0.0	0.0	0.0	1.0	1.0	1.0
1	0.0	0.0	0.0	1.0	1.0	1.0

df1.append(df2)

	0	1	2
0	0.0	0.0	0.0
1	0.0	0.0	0.0
0	1.0	1.0	1.0
1	1.0	1.0	1.0

导入数据

df = pd.read_csv('test.csv', header=None) #读取文件  # header没有columns
# df = pd.read_excel('test.excel',)
df.columns =['c1','c2','c3','c4']
df

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

df = df.dropna(thresh=4)
df

	c1	c2	c3	c4
1	4.9	3.0	1.4	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5

df.index = ['nick','jason','tank']

df.to_csv('test1.csv')

pandas基础中的基础,一定要学会, <奥卡姆剃刀>

df

	ttery	issue	code	code1	code2	time
0	min	20130801-3391	8,4,5,2,9	297734529	NaN	1013395466000
1	min	20130801-3390	7,8,2,1,2	298058212	NaN	1013395406000
2	min	20130801-3389	5,9,1,2,9	298329129	NaN	1013395346000
3	min	20130801-3388	3,8,7,3,3	298588733	NaN	1013395286000
4	min	20130801-3387	0,8,5,2,7	298818527	NaN	1013395226000

查看全文

相关阅读:
零点定理、介值定理
 古今数学史
 古今数学史
 StreamDM：基于Spark Streaming、支持在线学习的流式分析算法引擎
 StreamDM：基于Spark Streaming、支持在线学习的流式分析算法引擎
 随机化算法（二）
随机化算法（二）
机器学习算法时间复杂度的考虑
 机器学习算法时间复杂度的考虑
 matlab 工具函数、matlab toolbox（工具箱）

原文地址：https://www.cnblogs.com/aden668/p/11377896.html

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN

	c1	c2	c3	c4
0	5.1	NaN	1.4	0.2
1	4.9	3.0	1.4	0.2
2	4.7	3.2	NaN	0.2
3	7.0	3.2	4.7	1.4
4	6.4	3.2	4.5	1.5
5	6.9	3.1	4.9	NaN
6	NaN	NaN	NaN	NaN