3-13 索引进阶

 

1.指定倒序的索引

In [1]:

import pandas as pd
import numpy as np
s=pd.Series(np.arange(5),index=np.arange(5)[::-1],dtype='int64')#index=np.arange(5),指定索引范围；[::-1]是倒序索引
s

Out[1]:

4    0
3    1
2    2
1    3
0    4
dtype: int64

 

2.索引是否存在这些元素

In [2]:

s.isin([1,3,4])#索引是否存在这些元素

Out[2]:

4    False
3     True
2    False
1     True
0     True
dtype: bool

In [3]:

s[s.isin([1,3,4])]#取出对应元素

Out[3]:

3    1
1    3
0    4
dtype: int64

 

3.多重索引数据

In [4]:

s2=pd.Series(np.arange(6),index=pd.MultiIndex.from_product([[0,1],['a','b','c']]))#多重索引
s2

Out[4]:

0  a    0
   b    1
   c    2
1  a    3
   b    4
   c    5
dtype: int32

In [5]:

s2.iloc[s2.index.isin([(1,'a'),(0,'b')])]#多重索引

Out[5]:

0  b    1
1  a    3
dtype: int32

In [6]:

s

Out[6]:

4    0
3    1
2    2
1    3
0    4
dtype: int64

In [7]:

s[s>2]

Out[7]:

1    3
0    4
dtype: int64

 

4.索引数据

In [8]:

dates=pd.date_range('20181124',periods=8)#构造8天的数据
df=pd.DataFrame(np.random.randn(8,4),index=dates,columns=['A','B','C','D'])
df

Out[8]:

 

	A	B	C	D
2018-11-24	-1.302542	-1.850999	1.198842	0.134479
2018-11-25	-1.049921	0.277906	-1.470455	0.561132
2018-11-26	-0.572511	0.559077	-0.798659	1.680000
2018-11-27	0.163208	-0.775700	0.293595	-1.309184
2018-11-28	-0.341200	0.889370	0.278003	1.074634
2018-11-29	-1.399622	-0.577925	0.170549	0.872441
2018-11-30	1.146593	0.264301	-1.741237	-0.010080
2018-12-01	1.151403	-1.483099	0.425871	1.522704

In [9]:

df.select(lambda x:x=='A',axis='columns')#索引A列的数据

 

E:softwareAnaconda3 5.2.0libsite-packagesipykernel_launcher.py:1: FutureWarning: 'select' is deprecated and will be removed in a future release. You can use .loc[labels.map(crit)] as a replacement
  """Entry point for launching an IPython kernel.

Out[9]:

 

	A
2018-11-24	-1.302542
2018-11-25	-1.049921
2018-11-26	-0.572511
2018-11-27	0.163208
2018-11-28	-0.341200
2018-11-29	-1.399622
2018-11-30	1.146593
2018-12-01	1.151403

In [10]:

df['A']#和上面一样

Out[10]:

2018-11-24   -1.302542
2018-11-25   -1.049921
2018-11-26   -0.572511
2018-11-27    0.163208
2018-11-28   -0.341200
2018-11-29   -1.399622
2018-11-30    1.146593
2018-12-01    1.151403
Freq: D, Name: A, dtype: float64

 

5.where判断条件，并索引位置，替换数据

In [11]:

df.where(df<0)#把不满足条件的换成NaN

Out[11]:

 

	A	B	C	D
2018-11-24	-1.302542	-1.850999	NaN	NaN
2018-11-25	-1.049921	NaN	-1.470455	NaN
2018-11-26	-0.572511	NaN	-0.798659	NaN
2018-11-27	NaN	-0.775700	NaN	-1.309184
2018-11-28	-0.341200	NaN	NaN	NaN
2018-11-29	-1.399622	-0.577925	NaN	NaN
2018-11-30	NaN	NaN	-1.741237	-0.010080
2018-12-01	NaN	-1.483099	NaN	NaN

In [12]:

df.where(df<0,-df)#把不满足条件的换成-df,也就是可以替换成其他形式

Out[12]:

 

	A	B	C	D
2018-11-24	-1.302542	-1.850999	-1.198842	-0.134479
2018-11-25	-1.049921	-0.277906	-1.470455	-0.561132
2018-11-26	-0.572511	-0.559077	-0.798659	-1.680000
2018-11-27	-0.163208	-0.775700	-0.293595	-1.309184
2018-11-28	-0.341200	-0.889370	-0.278003	-1.074634
2018-11-29	-1.399622	-0.577925	-0.170549	-0.872441
2018-11-30	-1.146593	-0.264301	-1.741237	-0.010080
2018-12-01	-1.151403	-1.483099	-0.425871	-1.522704

 

6.query按条件查找索引

In [13]:

df=pd.DataFrame(np.random.rand(10,3),columns=list('abc'))
df

Out[13]:

 

	a	b	c
0	0.413798	0.636850	0.364235
1	0.338115	0.290723	0.413065
2	0.120505	0.151662	0.458175
3	0.032937	0.046397	0.231927
4	0.959934	0.210111	0.028029
5	0.319149	0.251683	0.409901
6	0.642514	0.613375	0.926091
7	0.777493	0.467535	0.576821
8	0.139102	0.978898	0.540589
9	0.543792	0.596241	0.330553

In [14]:

df.query('(a<b)')

Out[14]:

 

	a	b	c
0	0.413798	0.636850	0.364235
2	0.120505	0.151662	0.458175
3	0.032937	0.046397	0.231927
8	0.139102	0.978898	0.540589
9	0.543792	0.596241	0.330553

In [15]:

df.query('(a<b)&(b<c)')

Out[15]:

 

	a	b	c
2	0.120505	0.151662	0.458175
3	0.032937	0.046397	0.231927