查看dataframe的几种基础函数

import numpy as np
import pandas as pd

#日期作为索引，生成dataframe
dates = pd.date_range('20210821', periods=6)
dates

DatetimeIndex(['2021-08-21', '2021-08-22', '2021-08-23', '2021-08-24',
               '2021-08-25', '2021-08-26'],
              dtype='datetime64[ns]', freq='D')

df = pd.DataFrame(np.random.randn(6,4),index=dates,columns=list("ABCD"))#指定index以及columns
df

	A	B	C	D
2021-08-21	0.291046	-1.182064	-0.828510	-0.050328
2021-08-22	0.169817	0.372991	-0.009343	0.667243
2021-08-23	0.539149	-1.772332	-0.564575	-0.136001
2021-08-24	-0.696348	1.387040	0.387748	0.159869
2021-08-25	-0.756739	0.313842	0.673211	0.799936
2021-08-26	1.342560	0.528175	0.659946	1.509831

df.head(2)

	A	B	C	D
2021-08-21	-1.323688	1.019511	-0.338295	-3.620159
2021-08-22	-0.843637	0.453589	-0.991227	-0.405978

df.tail()

	A	B	C	D
2021-08-22	-0.843637	0.453589	-0.991227	-0.405978
2021-08-23	-0.554034	-2.888354	-0.666020	-0.337691
2021-08-24	0.067979	-1.051827	-0.415168	1.585267
2021-08-25	-0.673101	-0.837008	3.360230	0.980979
2021-08-26	-0.308061	-0.457008	2.063482	-0.477308

df.describe()

	A	B	C	D
count	6.000000	6.000000	6.000000	6.000000
mean	-0.605757	-0.626849	0.502167	-0.379149
std	0.473659	1.359844	1.774767	1.801394
min	-1.323688	-2.888354	-0.991227	-3.620159
25%	-0.801003	-0.998123	-0.603307	-0.459476
50%	-0.613567	-0.647008	-0.376731	-0.371835
75%	-0.369554	0.225940	1.463038	0.651311
max	0.067979	1.019511	3.360230	1.585267

df.index

DatetimeIndex(['2021-08-21', '2021-08-22', '2021-08-23', '2021-08-24',
               '2021-08-25', '2021-08-26'],
              dtype='datetime64[ns]', freq='D')

df.columns

Index(['A', 'B', 'C', 'D'], dtype='object')

df.index.dtype

dtype('<M8[ns]')

df.columns.dtype

dtype('O')

df.index[1] = "20211001"

---------------------------------------------------------------------------

TypeError                                 Traceback (most recent call last)

<ipython-input-13-a1c35869ebbd> in <module>
----> 1 df.index[1] = "20211001"


~/miniconda3/lib/python3.7/site-packages/pandas/core/indexes/base.py in __setitem__(self, key, value)
   3907 
   3908     def __setitem__(self, key, value):
-> 3909         raise TypeError("Index does not support mutable operations")
   3910 
   3911     def __getitem__(self, key):


TypeError: Index does not support mutable operations

df.columns[1]

'B'

df.columns[1] = "F"

---------------------------------------------------------------------------

TypeError                                 Traceback (most recent call last)

<ipython-input-15-e4a0ff29c1bd> in <module>
----> 1 df.columns[1] = "F"


~/miniconda3/lib/python3.7/site-packages/pandas/core/indexes/base.py in __setitem__(self, key, value)
   3907 
   3908     def __setitem__(self, key, value):
-> 3909         raise TypeError("Index does not support mutable operations")
   3910 
   3911     def __getitem__(self, key):


TypeError: Index does not support mutable operations

#我想改动index name，发现上述操作行不通，然后pandas提供了reindex接口
#To reindex means to conform the data to match a given set of labels along a particular axis.
#也就是说reindex相当于在指定轴上select，当没有的index出现则填充Nan
#我们再试一下会不会改动原来的dataframe
df.reindex(["C","A","E"],axis="columns")

	C	A	E
2021-08-21	-0.828510	0.291046	NaN
2021-08-22	-0.009343	0.169817	NaN
2021-08-23	-0.564575	0.539149	NaN
2021-08-24	0.387748	-0.696348	NaN
2021-08-25	0.673211	-0.756739	NaN
2021-08-26	0.659946	1.342560	NaN

df
#此时可以看到原来的dataframe并没有改变

	A	B	C	D
2021-08-21	0.291046	-1.182064	-0.828510	-0.050328
2021-08-22	0.169817	0.372991	-0.009343	0.667243
2021-08-23	0.539149	-1.772332	-0.564575	-0.136001
2021-08-24	-0.696348	1.387040	0.387748	0.159869
2021-08-25	-0.756739	0.313842	0.673211	0.799936
2021-08-26	1.342560	0.528175	0.659946	1.509831

df.reindex([0,5],axis="index") #此时使用数字索引好像失效了

	A	B	C	D
0	NaN	NaN	NaN	NaN
5	NaN	NaN	NaN	NaN

df.reindex(index=["2021.08.21","2021.08.26"],columns=["A","D"]) #index name 我打不出正确的格式，哈哈哈

	A	D
2021.08.21	NaN	NaN
2021.08.26	NaN	NaN

#骚操作1
#Note that the Index objects containing the actual axis labels can be shared between objects. So if we have a Series and a DataFrame
#rs = s.reindex(df.index)
#This means that the reindexed Series’s index is the same Python object as the DataFrame’s index.

#骚操作2
#reindex_like
df2 = pd.DataFrame(np.random.randn(4,3),index=["a","b","c","d"],columns=["A","B","C"])
df3 = pd.DataFrame(np.random.randn(4,4),index=["a","c","e","f"],columns=["A","C","E","F"])
print(df2)
print(df3)

          A         B         C
a -0.556821  1.481807  0.175237
b -0.440093  1.309766  0.265040
c -0.298950  0.565390  1.164518
d -0.287415 -0.226004  0.875915
          A         C         E         F
a -1.220393 -0.770392 -0.160964  0.344573
c  0.232881 -0.972749  1.882510  0.115421
e  0.265853 -1.628597 -1.128498 -0.073716
f -0.189951 -0.496491 -0.109640 -1.529581

df2.reindex_like(df3) #先取交集，再将模仿的数据框的其他行列拿过来填充NaN

	A	C	E	F
a	-0.556821	0.175237	NaN	NaN
c	-0.298950	1.164518	NaN	NaN
e	NaN	NaN	NaN	NaN
f	NaN	NaN	NaN	NaN

df3.reindex_like(df2)

	A	B	C
a	-1.220393	NaN	-0.770392
b	NaN	NaN	NaN
c	0.232881	NaN	-0.972749
d	NaN	NaN	NaN

#这里不再说reindex了，下一次深究一下
#我最初的目的是rename index/columns
df3

	A	C	E	F
a	-1.220393	-0.770392	-0.160964	0.344573
c	0.232881	-0.972749	1.882510	0.115421
e	0.265853	-1.628597	-1.128498	-0.073716
f	-0.189951	-0.496491	-0.109640	-1.529581

df3.rename(str.upper,axis="index") #df3.rename(str.upper,axis=0)

	A	C	E	F
A	-1.220393	-0.770392	-0.160964	0.344573
C	0.232881	-0.972749	1.882510	0.115421
E	0.265853	-1.628597	-1.128498	-0.073716
F	-0.189951	-0.496491	-0.109640	-1.529581

df3.rename({"a":"A","c":"B","e":"E","f":"F"},axis=0) #需要以字典的形式传递需要改动的标签
#实际的df3并未改动，设置inplace=True将会改动元数据

	A	C	E	F
A	-1.220393	-0.770392	-0.160964	0.344573
B	0.232881	-0.972749	1.882510	0.115421
E	0.265853	-1.628597	-1.128498	-0.073716
F	-0.189951	-0.496491	-0.109640	-1.529581

df3.rename({"a":"A"},axis=0)

	A	C	E	F
A	-1.220393	-0.770392	-0.160964	0.344573
c	0.232881	-0.972749	1.882510	0.115421
e	0.265853	-1.628597	-1.128498	-0.073716
f	-0.189951	-0.496491	-0.109640	-1.529581

df3.rename(index={},columns={}) #同样可行
df.columns = [x.lower() for x in df.columns] #其实也可以