Pandas DataFrame操作

DataFrame的创建

>>> import pandas as pd
>>> from pandas import DataFrame
#define a dict
>>> dic = {'Name':['Jeff','Lucy','Evan'],'Age':[28,26,27],'Sex':['Male','Female','Male']}
Load the dict to the dataframe
>>> df = DataFrame(dic)
>>> print df
   Age  Name     Sex
0   28  Jeff    Male
1   26  Lucy  Female
2   27  Evan    Male
#the order of the columns is default

#We define the order
>>> df1 = DataFrame(dic,columns=['Name','Sex','Age'])
>>> df1
   Name     Sex  Age
0  Jeff    Male   28
1  Lucy  Female   26
2  Evan    Male   27

#Define an empty column
>>> df1 = DataFrame(dic,columns=['Name','Age','Sex','Major'])
>>> df1
   Name  Age     Sex Major
0  Jeff   28    Male   NaN
1  Lucy   26  Female   NaN
2  Evan   27    Male   NaN

#Define the row name
>>> df1 = DataFrame(dic,columns=['Name','Age','Sex','Major'],index=['one','two','three'])
>>> df1
       Name  Age     Sex Major
one    Jeff   28    Male   NaN
two    Lucy   26  Female   NaN
three  Evan   27    Male   NaN

DataFrame内容读取与改变

>>> df1.columns
Index([u'Name', u'Age', u'Sex', u'Major'], dtype='object')
>>> df1.Sex
one        Male
two      Female
three      Male
Name: Sex, dtype: object

>>> df1['Sex']
one        Male
two      Female
three      Male
Name: Sex, dtype: object

>>> df1.ix['two']
Name       Lucy
Age          26
Sex      Female
Major       NaN
Name: two, dtype: object

>>> df1.index
Index([u'one', u'two', u'three'], dtype='object')

#Copy a colum from a Series
>>> df1
       Name  Age     Sex Major
one    Jeff   28    Male   NaN
two    Lucy   26  Female   NaN
three  Evan   27    Male   NaN
>>> s1 = (['Se','Se','Ce'])
>>> df1.Major=s1
>>> df1
       Name  Age     Sex Major
one    Jeff   28    Male    Se
two    Lucy   26  Female    Se
three  Evan   27    Male    Ce

#Define a new column
>>> df1['Type']=df1.Major=='Se'
>>> df1
       Name  Age     Sex Major   Type
one    Jeff   28    Male    Se   True
two    Lucy   26  Female    Se   True
three  Evan   27    Male    Ce  False

#Remove a column
>>> del df1['Type']
>>> df1
       Name  Age     Sex Major
one    Jeff   28    Male    Se
two    Lucy   26  Female    Se
three  Evan   27    Male    Ce

Other Methods to define

Define a DF with Two-layer Dict
>>> dic1={'name':{'1':'Jeff','2':'Mia','3':'Evan'},'age':{'1':28,'3':27,'2':18,'4':23}}
>>> df2=DataFrame(dic1)
>>> df2
   age  name
1   28  Jeff
2   18   Mia
3   27  Evan
4   23   NaN

Transpose
>>> df2.T
         1    2     3    4
age     28   18    27   23
name  Jeff  Mia  Evan  NaN

>>> df2.columns.name='items'
>>> df2.index.name='student_id'
>>> df2
items       age  name
student_id
1            28  Jeff
2            18   Mia
3            27  Evan
4            23   NaN

>>> df2.values
array([[28L, 'Jeff'],
       [18L, 'Mia'],
       [27L, 'Evan'],
       [23L, nan]], dtype=object)