stack,unstack,groupby,pivot_table的区别

stack,unstack,groupby,pivot_table的区别

stack() 堆积，是花括号的形式，只有列上的索引，unstack() 不要堆积，是表格的形式，行列均有索引 groupby() pivot_table 使用透视表实现groupby的功能

In [100]:

 

data=pd.DataFrame(np.arange(6).reshape((2,3)),index=pd.Index(['A','B']),columns=pd.Index(['one','two','three']))count()函数 列表推导式

#统计列表每个元素中指定单词出现的个数

words=['apple','pare','banana','and','peach','Anda']

for word in words:

    print(word.lower().count('a'))  #lower()识别大小写

1

1

3

1

1

2

[word for word in words if word.lower().count('a')>=2]

['banana', 'Anda']

strings=['a','bv','tit','apple','ctr']

[x.title() for x in strings if len(x)>2]

['Tit', 'Apple', 'Ctr']

list(map(len,strings))

[1, 2, 3, 5, 3]

transpose() 转置

import numpy as np

three=np.arange(18).reshape(2,3,3)

three

array([[[ 0,  1,  2],

        [ 3,  4,  5],

        [ 6,  7,  8]],

​

       [[ 9, 10, 11],

        [12, 13, 14],

        [15, 16, 17]]])

three.transpose(2,1,0)

array([[[ 0,  9],

        [ 3, 12],

        [ 6, 15]],

​

       [[ 1, 10],

        [ 4, 13],

        [ 7, 16]],

​

       [[ 2, 11],

        [ 5, 14],

        [ 8, 17]]])

​

 

 

In [101]:

data

 

 

Out[101]:

	one	two	three
A	0	1	2
B	3	4	5

In [106]:

 

dd=data.stack()

dd

 

 

Out[106]:

A  one      0
   two      1
   three    2
B  one      3
   two      4
   three    5
dtype: int64

In [107]:

dd.unstack()

 

 

Out[107]:

	one	two	three
A	0	1	2
B	3	4	5

In [112]:

 dd.unstack(level=0) #取最外层索引

 

 

Out[112]:

	A	B
one	0	3
two	1	4
three	2	5

In [113]:

 

dd.unstack(level=-1) #取内层索引

 

 

Out[113]:

	one	two	three
A	0	1	2
B	3	4	5

In [212]:

 

df = pd.DataFrame({'key1':['a','a','b','b','a'],'key2':['one','two','one','two','one'],'data1':np.random.randn(5),'data2':np.random.randn(5)})

In [115]:

df

 

 

Out[115]:

	key1	key2	data1	data2
0	a	one	-0.343458	-0.173529
1	a	two	-0.753353	0.068864
2	b	one	-0.554884	-0.147296
3	b	two	-0.064841	1.483495
4	a	one	0.237470	-0.107894

In [120]:

grouped1=df.groupby('key1')

 

 

In [121]:

 

grouped2=df.groupby(['key1','key2'])

 

 

In [122]:

[x for x in grouped1]

 

 

Out[122]:

[('a',   key1 key2     data1     data2
  0    a  one -0.343458 -0.173529
  1    a  two -0.753353  0.068864
  4    a  one  0.237470 -0.107894), ('b',   key1 key2     data1     data2
  2    b  one -0.554884 -0.147296
  3    b  two -0.064841  1.483495)]

In [123]:

[x for x in grouped2]

 

 

Out[123]:

[(('a', 'one'),   key1 key2     data1     data2
  0    a  one -0.343458 -0.173529
  4    a  one  0.237470 -0.107894),
 (('a', 'two'),   key1 key2     data1     data2
  1    a  two -0.753353  0.068864),
 (('b', 'one'),   key1 key2     data1     data2
  2    b  one -0.554884 -0.147296),
 (('b', 'two'),   key1 key2     data1     data2
  3    b  two -0.064841  1.483495)]

In [124]:

 

#pandas.pivot_table(data,values=None,index=None,columns=None,aggfunc='mean',fill_value=None,margins=False,\

                   #dropna=True,margins_name='All')[source]

#pivot_table 的默认函数是mean，即求平均值。

 

 

In [126]:

 

pd.pivot_table(df,index='key2',columns='key1')

 

 

Out[126]:

	data1		data2
key1	a	b	a	b
key2
one	-0.052994	-0.554884	-0.140712	-0.147296
two	-0.753353	-0.064841	0.068864	1.483495

In [127]:

 

pd.pivot_table(df,index=['key1','key2'])

 

 

Out[127]:

		data1	data2
key1	key2
a	one	-0.052994	-0.140712
	two	-0.753353	0.068864
b	one	-0.554884	-0.147296
	two	-0.064841	1.483495

In [130]:

 df.pivot_table('data1',columns='key2')

 

 

Out[130]:

key2	one	two
data1	-0.220291	-0.409097