数据清洗和准备

二、数据转换

移除重复数据

data = pd.DataFrame({'k1':['one','two']*3+['two'],
                     'k2':[1,1,2,3,3,4,4]})
data
Out:
    k1  k2
0  one   1
1  two   1
2  one   2
3  two   3
4  one   3
5  two   4
6  two   4

#检查 哪个重复  
data.duplicated()
Out:
0    False
1    False
2    False
3    False
4    False
5    False
6     True
dtype: bool

# 删除重复数据 
data.drop_duplicates()
Out:
    k1  k2
0  one   1
1  two   1
2  one   2
3  two   3
4  one   3
5  two   4

data['v1'] = range(7)
data
Out:
    k1  k2  v1
0  one   1   0
1  two   1   1
2  one   2   2
3  two   3   3
4  one   3   4
5  two   4   5
6  two   4   6

data.drop_duplicates(['k1'])  # 按照k1 这一列去除重复项
Out:
    k1  k2  v1
0  one   1   0
1  two   1   1

data.drop_duplicates(['k2'],keep='last')
Out:
    k1  k2  v1
1  two   1   1
2  one   2   2
4  one   3   4
6  two   4   6

data = pd.DataFrame({'k1':['one','two']*3+['two'],
                     'k2':[1,1,2,3,3,4,4]})
data
Out:
    k1  k2
0  one   1
1  two   1
2  one   2
3  two   3
4  one   3
5  two   4
6  two   4

data.drop_duplicates(keep='last')
Out:
    k1  k2
0  one   1
1  two   1
2  one   2
3  two   3
4  one   3
6  two   4

利用函数或映射进行数据转换

data = pd.DataFrame({'food': ['Apple', 'banana', 'orange','apple','Mango', 'tomato'],
                    'price': [4, 3, 3.5, 6, 12,3]})
data
Out:
     food  price
0   Apple    4.0
1  banana    3.0
2  orange    3.5
3   apple    6.0
4   Mango   12.0
5  tomato    3.0

meat = {'apple':'fruit',
       'banana':'fruit',
       'orange':'fruit',
       'mango':'fruit',
       'tomato':'vagetables'}

#值小写
low = data['food'].str.lower()
low
Out:
0     apple
1    banana
2    orange
3     apple
4     mango
5    tomato
Name: food, dtype: object

data['class'] = low.map(meat)
data
Out:
     food  price       class      class1
0   Apple    4.0       fruit       fruit
1  banana    3.0       fruit       fruit
2  orange    3.5       fruit       fruit
3   apple    6.0       fruit       fruit
4   Mango   12.0       fruit       fruit
5  tomato    3.0  vagetables  vagetables

data['class1'] = data['food'].map(lambda x:meat[x.lower()])
data
Out:
     food  price       class      class1
0   Apple    4.0       fruit       fruit
1  banana    3.0       fruit       fruit
2  orange    3.5       fruit       fruit
3   apple    6.0       fruit       fruit
4   Mango   12.0       fruit       fruit
5  tomato    3.0  vagetables  vagetables

data['class1'] = data['food'].map(lambda x: meat[x.lower()])
data
Out:
     food  price       class      class1
0   Apple    4.0       fruit       fruit
1  banana    3.0       fruit       fruit
2  orange    3.5       fruit       fruit
3   apple    6.0       fruit       fruit
4   Mango   12.0       fruit       fruit
5  tomato    3.0  vegetables  vegetables

替换值

data = pd.Series([1,-999,2,-1000,3])
data
Out:
0       1
1    -999
2       2
3   -1000
4       3
dtype: int64

data.replace(-999,np.nan)
Out:
0       1.0
1       NaN
2       2.0
3   -1000.0
4       3.0
dtype: float64

data.replace([-999,-1000],np.nan)  # 替换多个
Out:
0    1.0
1    NaN
2    2.0
3    NaN
4    3.0
dtype: float64

data1 = data.replace([-999,-1000],[np.nan,0])  # replace 会返回一个新的对象
data1
Out:
0    1.0
1    NaN
2    2.0
3    0.0
4    3.0
dtype: float64

data.replace({-999:np.nan,-1000:0})
Out:
0       1
1    -999
2       2
3   -1000
4       3
dtype: int64

重命名索引

data = pd.DataFrame(np.arange(12).reshape((3, 4)),
                    index=['BeiJing', 'Tokyo', 'New York'],
                    columns=['one', 'two', 'three', 'four'])
data
Out:
          one  two  three  four
BeiJing     0    1      2     3
Tokyo       4    5      6     7
New York    8    9     10    11

# 重新索引
data.reindex(['a', 'b', 'c'])  # reindex 只能修改已有的标签名
Out:
   one  two  three  four
a  NaN  NaN    NaN   NaN
b  NaN  NaN    NaN   NaN
c  NaN  NaN    NaN   NaN

data
Out:
          one  two  three  four
BeiJing     0    1      2     3
Tokyo       4    5      6     7
New York    8    9     10    11

#大写
tran = lambda x:x[:4].upper()  
data.index.map(tran)
Out:
Index(['BEIJ', 'TOKY', 'NEW '], dtype='object')

data.index = data.index.map(tran)
data
Out:
      one  two  three  four
BEIJ    0    1      2     3
TOKY    4    5      6     7
NEW     8    9     10    11

# rename
data.rename(index=str.title,columns = str.upper)
Out:
      ONE  TWO  THREE  FOUR
Beij    0    1      2     3
Toky    4    5      6     7
New     8    9     10    11

#结合字典型对象对标签更新
data.rename(index={'TOKY':'东京'},columns={'three':'第三年'})
Out:
      one  two  第三年  four
BEIJ    0    1    2     3
东京      4    5    6     7
NEW     8    9   10    11

data.rename(index={'TOKY':'东京'},columns={'three':'第三年'},inplace = True)
data
Out:
      one  two  第三年  four
BEIJ    0    1    2     3
东京      4    5    6     7
NEW     8    9   10    11