Pandas 基本技巧
1.数据查看和转置
import numpy as np
import pandas as pd
# 导入numpy、pandas模块 # 数据查看、转置 df = pd.DataFrame(np.random.rand(16).reshape(8,2)*100,
columns = ['a','b'])
print(df.head(2)) #查看前两条数据
print(df.tail())
# .head()查看头部数据
# .tail()查看尾部数据
# 默认查看5条 print(df.T)
# .T 转置
输出结果：
a b
0 64.231620 24.222954
1 3.004779 92.549576
a b
3 54.787062 17.264577
4 13.106864 5.500618
5 8.631310 79.109355
6 22.107241 94.901685
7 29.034599 54.156278
0 1 2 3 4 5 \
a 64.231620 3.004779 25.002825 54.787062 13.106864 8.631310
b 24.222954 92.549576 87.818090 17.264577 5.500618 79.109355 6 7
a 22.107241 29.034599
b 94.901685 54.156278
2.（1）添加与修改_1
# 添加与修改 df = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
columns = ['a','b','c','d'])
print(df) df['e'] = 10
df.loc[4] = 20
print(df)
# 新增列/行并赋值 df['e'] = 20
df[['a','c']] = 100
print(df)
# 索引后直接修改值 #注意：不能同时添加两列，否则会报错，如：df[['f','g']] = 200 ，必须一列一列的添加
输出结果：
a b c d
0 14.342082 52.604100 26.561995 60.441731
1 20.331108 43.537490 1.020098 7.171418
2 35.226542 9.573718 99.273254 0.867227
3 47.511549 56.783730 47.580639 67.007725
a b c d e
0 14.342082 52.604100 26.561995 60.441731 10
1 20.331108 43.537490 1.020098 7.171418 10
2 35.226542 9.573718 99.273254 0.867227 10
3 47.511549 56.783730 47.580639 67.007725 10
4 20.000000 20.000000 20.000000 20.000000 20
a b c d e
0 100 52.604100 100 60.441731 20
1 100 43.537490 100 7.171418 20
2 100 9.573718 100 0.867227 20
3 100 56.783730 100 67.007725 20
4 100 20.000000 100 20.000000 20
（2）添加与修改_2
import numpy as np
import pandas as pd df = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
columns = ['a','b','c','d'])
df.iloc[0] = 100
print(df)
df.iloc[0] = [1,2,3,4]
print(df) #增加一行尽量曲用loc去增加，iloc是不能增加的,会报错
df.loc[5] = 100
print(df)
输出结果：
a b c d
0 100.000000 100.000000 100.000000 100.000000
1 93.941010 7.951216 77.744847 66.842114
2 72.795874 40.031626 22.842638 92.876458
3 40.474858 53.663771 48.452597 66.444382
a b c d
0 1.000000 2.000000 3.000000 4.000000
1 93.941010 7.951216 77.744847 66.842114
2 72.795874 40.031626 22.842638 92.876458
3 40.474858 53.663771 48.452597 66.444382
a b c d
0 1.000000 2.000000 3.000000 4.000000
1 93.941010 7.951216 77.744847 66.842114
2 72.795874 40.031626 22.842638 92.876458
3 40.474858 53.663771 48.452597 66.444382
5 100.000000 100.000000 100.000000 100.000000
3.删除
（1）
# 删除 del / drop() df = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
columns = ['a','b','c','d'])
print(df) del df['a']
print(df)
print('-----')
# del语句 - 删除列
#注意：删除行的时候不能用del df.loc[index]或者df.iloc[index] 否则会报错 可以变相的删除 如删除第一行 可令df = df.iloc[1:] print(df.drop(0))
print(df.drop([1,2]))
print(df)
print('-----')
# drop()删除行，inplace=False → 删除后生成新的数据，不改变原数据 print(df.drop(['d'], axis = 1)) #axis =0 的时候删除行
print(df)
# drop()删除列，需要加上axis = 1，inplace=False → 删除后生成新的数据，不改变原数据
输出结果：
a b c d
0 71.238538 6.121303 77.988034 44.047009
1 34.018365 78.192855 50.467246 81.162337
2 86.311980 44.341469 49.789445 35.657665
3 78.073272 31.457479 74.385014 24.655976
b c d
0 6.121303 77.988034 44.047009
1 78.192855 50.467246 81.162337
2 44.341469 49.789445 35.657665
3 31.457479 74.385014 24.655976
-----
b c d
1 78.192855 50.467246 81.162337
2 44.341469 49.789445 35.657665
3 31.457479 74.385014 24.655976
b c d
0 6.121303 77.988034 44.047009
3 31.457479 74.385014 24.655976
b c d
0 6.121303 77.988034 44.047009
1 78.192855 50.467246 81.162337
2 44.341469 49.789445 35.657665
3 31.457479 74.385014 24.655976
-----
b c
0 6.121303 77.988034
1 78.192855 50.467246
2 44.341469 49.789445
3 31.457479 74.385014
b c d
0 6.121303 77.988034 44.047009
1 78.192855 50.467246 81.162337
2 44.341469 49.789445 35.657665
3 31.457479 74.385014 24.655976
（2）
import numpy as np
import pandas as pd df = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
columns = ['a','b','c','d'])
print(df.drop(0))
print(df) #源数据不会改变 print(df.drop(0,inplace = True)) #这个方法改变了源数据，并不生成新的值了，所以输出为空
print(df) #有inplace 参数的时候就替换了源数据
输出结果：
a b c d
1 78.187118 19.237655 94.443127 67.466532
2 37.921956 84.157197 23.311418 24.128222
3 12.330334 6.034799 62.023747 28.034041
a b c d
0 60.558857 94.367826 88.690379 33.957380
1 78.187118 19.237655 94.443127 67.466532
2 37.921956 84.157197 23.311418 24.128222
3 12.330334 6.034799 62.023747 28.034041
None
a b c d
1 78.187118 19.237655 94.443127 67.466532
2 37.921956 84.157197 23.311418 24.128222
3 12.330334 6.034799 62.023747 28.034041
4.对齐
# 对齐 df1 = pd.DataFrame(np.random.randn(10, 4), columns=['A', 'B', 'C', 'D'])
df2 = pd.DataFrame(np.random.randn(7, 3), columns=['A', 'B', 'C'])
print(df1)
print(df2)
print(df1 + df2) #有共同的列名和共同的标签的话 就会相加 。没有共同的部分就会变为空值。任何值和空值进行运算都会变为空值
# DataFrame对象之间的数据自动按照列和索引（行标签）对齐 ,
输出结果：
A B C D
0 -1.528903 0.519125 -0.214881 -0.591775
1 -0.334501 -0.837666 0.568927 -0.599237
2 0.753145 0.569262 -1.181976 1.225363
3 -0.177136 -0.367530 0.382826 1.447591
4 0.215967 -0.612947 0.844906 0.130414
5 0.414375 -0.207225 0.140776 1.086686
6 0.008855 2.873956 -0.650806 -2.631485
7 -0.634085 0.625107 0.046198 -0.352343
8 0.646812 0.928476 0.519168 -0.644997
9 -0.697006 -0.178875 0.856392 -0.512101
A B C
0 -0.373297 0.607873 0.120016
1 0.343563 -2.901778 -0.370051
2 0.428568 0.319359 -3.263585
3 1.042845 -0.314763 -0.198816
4 0.071258 -0.484855 0.563127
5 -2.270312 -0.145558 0.931203
6 2.493652 -0.232491 -0.216451
A B C D
0 -1.902200 1.126998 -0.094865 NaN
1 0.009061 -3.739444 0.198876 NaN
2 1.181713 0.888620 -4.445561 NaN
3 0.865710 -0.682293 0.184010 NaN
4 0.287224 -1.097802 1.408034 NaN
5 -1.855938 -0.352783 1.071979 NaN
6 2.502507 2.641465 -0.867257 NaN
7 NaN NaN NaN NaN
8 NaN NaN NaN NaN
9 NaN NaN NaN NaN
6.排序
（1）按值排序
# 排序1 - 按值排序 .sort_values
# 同样适用于Series df1 = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
columns = ['a','b','c','d'])
print(df1)
print(df1.sort_values(['a'], ascending = True)) # 升序
#也可以这样写：print(df1.sort_values(by = 'a',ascending = True))
print(df1.sort_values(['a'], ascending = False)) # 降序
print('------')
# ascending参数：设置升序降序，默认升序
# 单列排序 df2 = pd.DataFrame({'a':[1,1,1,1,2,2,2,2],
'b':list(range(8)),
'c':list(range(8,0,-1))})
print(df2)
print(df2.sort_values(['a','c']))
# 多列排序，按列顺序排序
# 注意inplace参数
输出结果：
a b c d
0 28.598118 8.037050 51.856085 45.859414
1 91.412263 59.797819 27.912198 6.996883
2 92.001255 76.467245 76.524894 33.463836
3 47.054750 37.376781 94.286800 53.429360
a b c d
0 28.598118 8.037050 51.856085 45.859414
3 47.054750 37.376781 94.286800 53.429360
1 91.412263 59.797819 27.912198 6.996883
2 92.001255 76.467245 76.524894 33.463836
a b c d
2 92.001255 76.467245 76.524894 33.463836
1 91.412263 59.797819 27.912198 6.996883
3 47.054750 37.376781 94.286800 53.429360
0 28.598118 8.037050 51.856085 45.859414
------
a b c
0 1 0 8
1 1 1 7
2 1 2 6
3 1 3 5
4 2 4 4
5 2 5 3
6 2 6 2
7 2 7 1
a b c
3 1 3 5
2 1 2 6
1 1 1 7
0 1 0 8
7 2 7 1
6 2 6 2
5 2 5 3
4 2 4 4
（2）索引排序
# 排序2 - 索引排序 .sort_index df1 = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
index = [5,4,3,2],
columns = ['a','b','c','d'])
df2 = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
index = ['h','s','x','g'],
columns = ['a','b','c','d'])
print(df1)
print(df1.sort_index())
print(df2)
print(df2.sort_index())
# 按照index排序
# 默认 ascending=True, inplace=False
输出结果：
a b c d
5 80.932585 71.991854 64.582943 23.443231
4 82.054030 87.459058 12.108433 83.047490
3 56.329863 14.926822 47.884418 59.880352
2 0.347007 69.794103 74.375345 12.736429
a b c d
2 0.347007 69.794103 74.375345 12.736429
3 56.329863 14.926822 47.884418 59.880352
4 82.054030 87.459058 12.108433 83.047490
5 80.932585 71.991854 64.582943 23.443231
a b c d
h 53.041921 93.834097 13.423132 82.702020
s 0.003814 75.721426 73.086606 20.597472
x 32.678307 58.369155 70.487505 24.833117
g 46.232889 19.365147 9.872537 98.246438
a b c d
g 46.232889 19.365147 9.872537 98.246438
h 53.041921 93.834097 13.423132 82.702020
s 0.003814 75.721426 73.086606 20.597472
x 32.678307 58.369155 70.487505 24.833117
（3）
df1 = pd.DataFrame(np.random.rand(16).reshape(4,4)*100,
index = [5,4,3,2],
columns = ['a','b','c','d'])
print(df1)
print(df1.sort_index())
print(df1) # df1并没有变 print(df1.sort_index(inplace = True))
print(df1) # df1发生改变
输出结果：
a b c d
5 45.004735 23.449962 52.756124 60.237141
4 74.945903 63.813663 29.937821 66.420415
3 45.737208 82.376775 80.615108 40.479094
2 41.743173 82.013411 83.372130 76.195150
a b c d
2 41.743173 82.013411 83.372130 76.195150
3 45.737208 82.376775 80.615108 40.479094
4 74.945903 63.813663 29.937821 66.420415
5 45.004735 23.449962 52.756124 60.237141
a b c d
5 45.004735 23.449962 52.756124 60.237141
4 74.945903 63.813663 29.937821 66.420415
3 45.737208 82.376775 80.615108 40.479094
2 41.743173 82.013411 83.372130 76.195150
None
a b c d
2 41.743173 82.013411 83.372130 76.195150
3 45.737208 82.376775 80.615108 40.479094
4 74.945903 63.813663 29.937821 66.420415
5 45.004735 23.449962 52.756124 60.237141
练习：
作业1：创建一个3*3，值在0-100区间随机值的Dataframe（如图），分别按照index和第二列值大小，降序排序
import numpy as np
import pandas as pd
#练习1
# df = pd.DataFrame(np.random.rand(9).reshape(3,3)*100,
# index=['a','b','c'],
# columns=['v1','v2','v3'])
# print(df)
#
# print(df.sort_index())
# df.sort_values(by = 'v2',ascending= False,inplace = True)
# print(df)
作业2：创建一个5*2，值在0-100区间随机值的Dataframe（如图）df1，通过修改得到df2
#练习2
# df1 = pd.DataFrame(np.random.rand(10).reshape(5,2)*100,
# index=['a','b','c','d','e'],
# columns=['v1','v2'])
# print(df1)
# print(df1.drop(['e'],axis = 0).T)
作业3：如图创建Series，并按照要求修改得到结果
#练习3
df2 = pd.Series(np.arange(10),index= ['a','b','c','d','e','f','g','h','i','j'])
print(df2)
df2.loc[['a','e','f']] = 100
print(df2)
#或者
# df2.iloc[0] = 100
# df2.iloc[3] = 100
# df2.iloc[4] = 100
Pandas 基本技巧的更多相关文章
- pandas 使用技巧
多列展示 查看DataFrame时,列数较多会省略中间列的信息,可以设置最大展示列数,代码如下: 1 import pandas as pd 2 pd.set_option('display.max_ ...
- pandas 常用技巧总结
切片: loc:df.loc[num]:选择df 某一行 seriesdf.loc[[num1,num2]]: 选择df 某几行df.loc[[True,False,True, ,True]]: ...
- pandas 小技巧
1.找出某个字段包含某字符串的行: my_df[my_df['col_B'].str.contains('大连') > 0]或者 my_df[my_df['col_B'].apply(lambd ...
- pandas取dataframe特定行/列
1. 按列取.按索引/行取.按特定行列取 import numpy as np from pandas import DataFrame import pandas as pd df=DataFram ...
- 6个冷门但实用的pandas知识点
1 简介 pandas作为开展数据分析的利器,蕴含了与数据处理相关的丰富多样的API,使得我们可以灵活方便地对数据进行各种加工,但很多pandas中的实用方法其实大部分人都是不知道的,今天就来给大家介 ...
- 多快好省地使用pandas分析大型数据集
1 简介 pandas虽然是个非常流行的数据分析利器,但很多朋友在使用pandas处理较大规模的数据集的时候经常会反映pandas运算"慢",且内存开销"大". ...
- 数据处理：12个使得效率倍增的pandas技巧
数据处理:12个使得效率倍增的pandas技巧 1. 背景描述 Python正迅速成为数据科学家偏爱的语言,这合情合理.它拥有作为一种编程语言广阔的生态环境以及众多优秀的科学计算库.如果你刚开始学习P ...
- python/numpy/pandas数据操作知识与技巧
pandas针对dataframe各种操作技巧集合: filtering: 一般地,使用df.column > xx将会产生一个只有boolean值的series,以该series作为dataf ...
- Pandas一些小技巧
Pandas有一些不频繁使用容易忘记的小技巧 1.将不同Dataframe写在一个Excel的不同Sheet,或添加到已有Excel的不同Sheet(同名Sheet会覆盖) from pandas i ...
随机推荐
- div的浮动(float)
什么是浮动 浮动,故名思议,就是移动位置. 之所以不叫移动,而叫浮动,那是因为给元素设置浮动后,元素会浮到文档上面来,术语叫脱离文档流. 例子 下面我们看例子 <html> <hea ...
- django 定制admin
https://www.cnblogs.com/liwenzhou/p/9519321.html
- 动软代码生成器，主子表增加的时候子表的parentID无法插入问题解决方案
StringBuilder strSql=new StringBuilder(); strSql.Append("insert into HT_XunJiaMain("); str ...
- 探索Skip List (跳跃表)
附William Pugh的论文 Skip Lists: A Probabilistic Alternative to Balanced Trees 写在前面 以下内容针对的是Skip List的插入 ...
- PHP : 封装Mysqli的连接、关闭和增改查(面向过程)
废话不多说,直接上图和案例:(看行数进行拼接) 注意:连接数据库中,因为用了$CONNECT全局变量,所以我们的连接数据库封装方法必须要执行了才能获取到此全局变量的值 所引入的数据库文件(databa ...
- Last_SQL_Errno: 1050
主库上create table,从库上存在. 报错信息如下所示: Last_SQL_Errno: 1050 Last_SQL_Error: ...
- Gym 100090D Insomnia
从 n 变到 1,有多少种方案? 打表记忆化. #include <bits/stdc++.h> using namespace std; int n; ]; int dfs(int n) ...
- 在vue中使用插槽 slot
插槽(slot)这个概念非常重要 插槽的使用场景1:在子组件里面显示父组件的dom <div id='root'> <child content = '<p>Dell&l ...
- 【转】startActivityForResult和setResult详解
startActivityForResult与startActivity的不同之处在于:1.startActivity( ) 仅仅是跳转到目标页面,若是想跳回当前页面,则必须再使用一次startAct ...
- python 最简单的web应用(一)
对于所有的Web应用,本质上其实就是一个socket服务端,用户的浏览器其实就是一个socket客户端. server.py文件 #!/usr/bin/env python # -*- coding: ...