panda库的简易学习

布满荆棘的人生 2022-05-18 00:19 223阅读 0赞

## 1 重新生成索引 如果某个索引值不存在就引入缺失值 ##

1 from pandas import Series,DataFrame
    2 import pandas as pd
    3 import numpy as np
    4 obj=Series([4.5,7.2,-5.3,3.6],index=['d','b','a','c'])
    5 obj
    6 
    7 #重新生成索引
    8 obj2=obj.reindex(['a','b','c','d','e'])
    9 obj2

![936260-20170821102204761-884840345.png][]

![936260-20170821102223558-911603909.png][]

## a使用method的ffill可以实现前向值填充，效果如下 ##

1 #前向填充
    2 obj3=Series(['blue','purple','yellow'],index=[0,2,4])
    3 obj3.reindex(range(6),method='ffill')

![936260-20170821102614058-1992126259.png][]

![936260-20170821102650980-1370441654.png][]

## b:对于dataframe使用reindex可以同时修改行列索引，如果仅传入一个序列那么如下 ##

1 frame=DataFrame(np.arange(9).reshape((3,3)),index=['a','c','d'],
    2                 columns=['ohio','Texas','california'])
    3 frame

![936260-20170821103142105-2096442697.png][]

1 frame2=frame.reindex(['a','b','c','d'])
    2 frame2

![936260-20170821103216808-2007574735.png][]

## c：使用colunms重新索引列 ##

1 states=['Texax','Utah','california']
    2 frame.reindex(columns=states)

![936260-20170821103358793-1495549229.png][]

## 　d:同时插入行列，但是插值只能按行应用 ##

1 #同时对行 列进行重新索引 而插值只能引用到行
    2 frame.reindex(index=['a','b','c','d'],method='ffill',
    3               columns=states)

![936260-20170821103739621-2066003673.png][]

reindex的参数说明如下：

![936260-20170821103824277-1633876119.png][]

## 2 丢弃制定轴上的项 ##

## 　a:drop方法返回一个指定轴上删除了指定值的新对象，删除行c ##

1 #丢弃指定轴的项
    2 obj=Series(np.arange(5.),index=['a','b','c','d','e'])
    3 new_obj=obj.drop('c')
    4 new_obj

![936260-20170821104116668-891884480.png][]

## b：删除两个 b c ##

obj.drop(['d','c'])

![936260-20170821104250886-1721457521.png][]

c：对于dataframe可以删除任意轴上的索引

1 #对于DataFrame可以删除任意轴的索引
    2 data = DataFrame(np.arange(16).reshape((4,4)),
    3                  index=['ohio','colorado','utah','new york'],
    4                    columns=['one','two','three','four'])
    5 #删除两个
    6 data.drop(['colorado','ohio'])

![936260-20170821104427105-1462849673.png][]

## 3 索引，选取和过滤 ##

a:Series中的索引类似与Numpy，但是不只是整数，索引字符

1 obj=Series(np.arange(4.),index=['a','b','c','d'])
    2 obj['b']#1.0

![936260-20170821104652964-801787640.png][]

![936260-20170821104705855-1111479121.png][]

b：按照整数，范围

1 obj[1]#1.0
    2 obj[2:4]# 2 3

![936260-20170821104819121-1719754099.png][]

## c：利用标签的切片运算和普通depython切片不同，其包含末端 ##

1 obj['b':'c']#b c 1 2

![936260-20170821105008777-1839434533.png][]

## d：那么对dataframe进行索引就是获取一个或者多个列 ##

1 data=DataFrame(np.arange(16).reshape(4,4),
    2            index=['ohio','colorado','mike','jason'],
    3                  columns=['one','two','three','four'])
    4 data

![936260-20170821105139339-1294227135.png][]

## e：选择一列 ##

data['two']#输出第二列+行号 也就是索引

![936260-20170821105228871-265996361.png][]

## f：选择多列 ##

1 data[['three','one']]

![936260-20170821105306136-1008408825.png][]

## g：选取行标签前两行 ##

data[:2]#选取的是前面两行

![936260-20170821105425214-218133556.png][]

## h：选取第三列大于5的值 ##

data[data['three']>5]

![936260-20170821105524105-1323821243.png][]

## i：为了能在dataframe的行上进行标签索引引入字段ix ##

data.ix['colorado',['two','three']]

![936260-20170821105734621-2041733000.png][]

## j：选取第4 1 2列 而且行为colorado jason ##

data.ix[['colorado','jason'],[3,0,1]]

![936260-20170821110416605-1365775537.png][]

k：输出行mike

data.ix[2]

![936260-20170821110547871-1235335450.png][]

DataFrame索引总结

![936260-20170821110710308-289231813.png][]

## 4 算数运算和数据对齐 ##

a:Series的加法

1 s1=Series([7.3,-2.5,3.4,1.5],index=['a','c','d','e'])
    2 s2=Series([-2.1,3.6,-1.5,4,3.1],index=['a','c','e','f','g'])
    3 s1+s2

![936260-20170821111927886-1410575191.png][]![936260-20170821112012480-849176936.png][]

![936260-20170821112036480-1854599995.png][]

## 　b:对于dataframe，对齐会同时发生在行 列中 ##

df1=DataFrame(np.arange(9.).reshape((3,3)),columns=list('bcd'),
                  index=['utah','ohio','colorado'])
    df2=DataFrame(np.arange(12.).reshape((4,3)),columns=list('bde'),
                  index=['utah','ohio','colorado','oragen'])

![936260-20170821112655386-2029138938.png][]

![936260-20170821112710839-1827860531.png][]

df1+df2

![936260-20170821112751011-681684908.png][]

## \------>索引和列都为其并集 ##

c：在算术方法中填充值。比如说两个dataframe相加，其中一个不在的时候填充为0

1 #算术中进行填充
    2 df1=DataFrame(np.arange(12.).reshape((3,4)),columns=list('abcd'))
    3 df2=DataFrame(np.arange(20.).reshape((4,5)),columns=list('abcde'))
    4 df1+df2

![936260-20170821113435496-877897963.png][]

#使用df1的add方法 传入df2以及一个fill_value参数
    df1.add(df2,fill_value=0)

![936260-20170821113618355-198820315.png][]

5 DataFrame和Series之间的运算----->广播，也就是如果第一个数值-1，那么这个列都会减1

a：看一看一个二维数组和一行之间的差

arr=np.arange(12.).reshape((3,4))

![936260-20170821114305293-1620393111.png][]

arr[0]

![936260-20170821114410449-1299708908.png][]

arr-arr[0]

![936260-20170821114449027-543625290.png][]

b:frame和series的运算

frame=DataFrame(np.arange(12.).reshape((4,3)),columns=list('bde'),
                    index=['utah','ohio','texas','orogen'])
    series=frame.ix[0]

![936260-20170821114904339-25156645.png][]

![936260-20170821114920839-139030623.png][]

frame-series

![936260-20170821114956199-710869065.png][]

[936260-20170821102204761-884840345.png]: /images/20220518/8e0167c1a6e4404f87b0059d892b87dd.png
[936260-20170821102223558-911603909.png]: /images/20220518/c78eac0545374fd59accfb5a3c68623a.png
[936260-20170821102614058-1992126259.png]: /images/20220518/74e95fb9d9da4a5a88e912fb9e4e511b.png
[936260-20170821102650980-1370441654.png]: /images/20220518/c068ae7191274feb9791061984c7d51d.png
[936260-20170821103142105-2096442697.png]: /images/20220518/9fe97f918a004e0fb32b566b4e6f7739.png
[936260-20170821103216808-2007574735.png]: /images/20220518/df767fde3d0e44d388fc1f299f39e287.png
[936260-20170821103358793-1495549229.png]: /images/20220518/4265072c2e354a709b11aa1569d4d1b0.png
[936260-20170821103739621-2066003673.png]: /images/20220518/382597aa1f8d4b29a26867c71a65334b.png
[936260-20170821103824277-1633876119.png]: /images/20220518/6363c80b8dba4bdbbef6c5f79f4fbbd3.png
[936260-20170821104116668-891884480.png]: /images/20220518/a4601c5401224be5b183d183ad90ecf1.png
[936260-20170821104250886-1721457521.png]: /images/20220518/ffc0df8c069f4c73a6adfa5f20044556.png
[936260-20170821104427105-1462849673.png]: /images/20220518/e6868711c4f54c51ab4ab78446682eac.png
[936260-20170821104652964-801787640.png]: /images/20220518/3e7c44ea80784bcf995216f8b35cbe1c.png
[936260-20170821104705855-1111479121.png]: /images/20220518/d571881c0a094c2aa6a6d39e7b2c0710.png
[936260-20170821104819121-1719754099.png]: /images/20220518/844c6445b56948b3a5d7849281ef31c7.png
[936260-20170821105008777-1839434533.png]: /images/20220518/a71a6ffb5f8749808d7f981f88dccf8a.png
[936260-20170821105139339-1294227135.png]: /images/20220518/5d1c344412d74a458e3a8054d7447414.png
[936260-20170821105228871-265996361.png]: /images/20220518/e5e702f6bf7b4baa9d2426cd87ee3510.png
[936260-20170821105306136-1008408825.png]: /images/20220518/1ed6ca0e06094586ab0048ef2fe07f2b.png
[936260-20170821105425214-218133556.png]: /images/20220518/57bc50c949ab4c03bafbeddf593e42f1.png
[936260-20170821105524105-1323821243.png]: /images/20220518/61a8b075a2154e0ea13f1fcd5c85c9ff.png
[936260-20170821105734621-2041733000.png]: /images/20220518/a8db9273cf3c4f0398f59f6de13d39bd.png
[936260-20170821110416605-1365775537.png]: /images/20220518/b598f375b6fa400fb84271455a627f39.png
[936260-20170821110547871-1235335450.png]: /images/20220518/4ae7406fc2e448e7b66a02a679a0fe8b.png
[936260-20170821110710308-289231813.png]: /images/20220518/56f8f9b2dcdc4edb9c67c0f177d8304f.png
[936260-20170821111927886-1410575191.png]: https://images2017.cnblogs.com/blog/936260/201708/936260-20170821111927886-1410575191.png
[936260-20170821112012480-849176936.png]: /images/20220518/21e5ca6073274dbd94e39dc620de7b33.png
[936260-20170821112036480-1854599995.png]: /images/20220518/1de94a0b4d9c4525b5e2435bd13a1fb9.png
[936260-20170821112655386-2029138938.png]: /images/20220518/fa7897b18e09473483fd570f4a82e3fb.png
[936260-20170821112710839-1827860531.png]: /images/20220518/ef01f9f0158d4bd7a6fedb1ab05d5ddd.png
[936260-20170821112751011-681684908.png]: /images/20220518/c754bc7987e04125abfb9f94f3dabb4a.png
[936260-20170821113435496-877897963.png]: /images/20220518/e0f5e789c6094af1b84e9cc853f96730.png
[936260-20170821113618355-198820315.png]: /images/20220518/04f63ba8f5314ae098102cea68b6a064.png
[936260-20170821114305293-1620393111.png]: /images/20220518/bf12347af5bc4e628d962e2c2b0acae8.png
[936260-20170821114410449-1299708908.png]: /images/20220518/31d23b5259d7477991db28f8df3b98b5.png
[936260-20170821114449027-543625290.png]: /images/20220518/1c2fcca474cd4ca5a2b8fe58806842c4.png
[936260-20170821114904339-25156645.png]: /images/20220518/95f44d09018146d28accf1ae98b3bd01.png
[936260-20170821114920839-139030623.png]: /images/20220518/81a0781d0e424d5495f6322f07b385f3.png
[936260-20170821114956199-710869065.png]: /images/20220518/a387b48740864ed3935f7a880b47a3c4.png