DAY 8 标签编码与连续变量处理

DAY 8 标签编码与连续变量处理

1.字典的简单介绍

# 使用花括号创建字典
dict={"name":"Alice","age":25,"city":"New York"}
dict

{'name': 'Alice', 'age': 25, 'city': 'New York'}

# 访问字典中的值
dict["name"]

'Alice'

2.标签编码

离散特征：

不存在顺序，采用 get_dummies() 函数进行独热编码；

存在顺序和大小关系，采用 map() 函数进行标签编码。

# 读取数据
import pandas as pd
data=pd.read_csv(r"data.csv")
data.head()

	Id	Home Ownership	Annual Income	Years in current job	Tax Liens	Number of Open Accounts	Years of Credit History	Maximum Open Credit	Number of Credit Problems	Months since last delinquent	Bankruptcies	Purpose	Term	Current Loan Amount	Current Credit Balance	Monthly Debt	Credit Score	Credit Default
0	0	Own Home	482087.0	NaN	0.0	11.0	26.3	685960.0	1.0	NaN	1.0	debt consolidation	Short Term	99999999.0	47386.0	7914.0	749.0	0
1	1	Own Home	1025487.0	10+ years	0.0	15.0	15.3	1181730.0	0.0	NaN	0.0	debt consolidation	Long Term	264968.0	394972.0	18373.0	737.0	1
2	2	Home Mortgage	751412.0	8 years	0.0	11.0	35.0	1182434.0	0.0	NaN	0.0	debt consolidation	Short Term	99999999.0	308389.0	13651.0	742.0	0
3	3	Own Home	805068.0	6 years	0.0	8.0	22.5	147400.0	1.0	NaN	1.0	debt consolidation	Short Term	121396.0	95855.0	11338.0	694.0	0
4	4	Rent	776264.0	8 years	0.0	13.0	13.6	385836.0	1.0	NaN	0.0	debt consolidation	Short Term	125840.0	93309.0	7180.0	719.0	0

Home Ownership 特征

data["Home Ownership"].value_counts()

Home Ownership
Home Mortgage    3637
Rent             3204
Own Home          647
Have Mortgage      12
Name: count, dtype: int64

自定义抗风险能力：Own Home < Rent < Have Mortgage < Home Mortgage

# 定义映射字典
mapping={"Own Home":0,"Rent":1,"Have Mortgage":2,"Home Mortgage":3}
data["Home Ownership"].head()

0         Own Home
1         Own Home
2    Home Mortgage
3         Own Home
4             Rent
Name: Home Ownership, dtype: object

# 标签编码
data["Home Ownership"]=data["Home Ownership"].map(mapping)
data["Home Ownership"].head()

0    0
1    0
2    3
3    0
4    1
Name: Home Ownership, dtype: int64

Term 特征

data["Term"].value_counts()

Term
Short Term    5556
Long Term     1944
Name: count, dtype: int64

# 定义映射字典
mapping={"Short Term":1,"Long Term":0}

# 标签编码
data["Term"]=data["Term"].map(mapping)
data["Term"].head()

0    1
1    0
2    1
3    1
4    1
Name: Term, dtype: int64

一个映射函数实现两次编码

# 重新读取数据
import pandas as pd
data=pd.read_csv(r"data.csv")

mapping={
    "Home Ownership":{"Own Home":0,"Rent":1,"Have Mortgage":2,"Home Mortgage":3},
    "Term":{"Short Term":1,"Long Term":0}
}

# 对 Home Ownership 列进行映射
data["Home Ownership"]=data["Home Ownership"].map(mapping["Home Ownership"])
data["Home Ownership"].head()

0    0
1    0
2    3
3    0
4    1
Name: Home Ownership, dtype: int64

# 对 Term 列进行映射
data["Term"]=data["Term"].map(mapping["Term"])
data["Term"].head()

0    1
1    0
2    1
3    1
4    1
Name: Term, dtype: int64

3.连续特征的处理：归一化和标准化

Annual Income 特征

# 归一化处理函数 manual_normalize
def manual_normalize(data):
    min_val=data.min()
    max_val=data.max()
    normalized_data=(data-min_val)/(max_val-min_val)
    return normalized_data

data["Annual Income"]=manual_normalize(data["Annual Income"])
data["Annual Income"].head()

0    0.031798
1    0.086221
2    0.058771
3    0.064145
4    0.061260
Name: Annual Income, dtype: float64

# 使用 sklearn 库进行归一化处理
from sklearn.preprocessing import MinMaxScaler

# 重新读取数据
import pandas as pd
data=pd.read_csv(r"data.csv")

min_max_scaler=MinMaxScaler()
data["Annual Income"]=min_max_scaler.fit_transform(data[["Annual Income"]])
data["Annual Income"].head()

0    0.031798
1    0.086221
2    0.058771
3    0.064145
4    0.061260
Name: Annual Income, dtype: float64

# 标准化处理
from sklearn.preprocessing import StandardScaler

# 重新读取数据
import pandas as pd
data=pd.read_csv(r"data.csv")

scaler=StandardScaler()
data["Annual Income"]=scaler.fit_transform(data[["Annual Income"]])
data["Annual Income"].head()

0   -1.046183
1   -0.403310
2   -0.727556
3   -0.664078
4   -0.698155
Name: Annual Income, dtype: float64

作业：对心脏病数据集的特征用上述知识完成，一次性用所有的处理方式完成预处理，尝试手动完成，多敲几遍代码。

离散特征的处理：标签编码

# 读取数据
import pandas as pd
data=pd.read_csv(r"heart.csv")

# 离散特征
list_discrete=["sex","cp","fbs","restecg","exang","slope","thal"]

# 使用 astype 指定数据类型
for i in list_discrete:
    data[i]=data[i].astype("object")

data.dtypes

age           int64
sex          object
cp           object
trestbps      int64
chol          int64
fbs          object
restecg      object
thalach       int64
exang        object
oldpeak     float64
slope        object
ca            int64
thal         object
target        int64
dtype: object

# 定义映射字典
mapping={"sex":{0:0,1:1},"cp":{0:1,1:2,2:3,3:4},"fbs":{0:0,1:1},"restecg":{0:0,1:1,2:2},"exang":{0:0,1:1},"slope":{0:1,1:2,2:3},"thal":{0:1,1:2,2:3,3:4}}

# 标签编码
for i in list_discrete:
    data[i]=data[i].map(mapping[i])

data.head()

	age	sex	cp	trestbps	chol	fbs	restecg	thalach	exang	oldpeak	slope	ca	thal	target
0	63	1	4	145	233	1	0	150	0	2.3	1	0	2	1
1	37	1	3	130	250	0	1	187	0	3.5	1	0	3	1
2	41	0	2	130	204	0	0	172	0	1.4	3	0	3	1
3	56	1	2	120	236	0	1	178	0	0.8	3	0	3	1
4	57	0	1	120	354	0	1	163	1	0.6	3	0	3	1

连续特征的处理：归一化和标准化

# 读取数据
import pandas as pd
data=pd.read_csv(r"heart.csv")

# 连续特征
list_continuous=["age","trestbps","chol","thalach","oldpeak","ca"]

# 归一化处理函数 manual_normalize
def manual_normalize(data):
    min_val=data.min()
    max_val=data.max()
    normalized_data=(data-min_val)/(max_val-min_val)
    return normalized_data

# 归一化
for i in list_continuous:
    data[i]=manual_normalize(data[i])

data.head()

	age	sex	cp	trestbps	chol	fbs	restecg	thalach	exang	oldpeak	slope	ca	thal	target
0	0.708333	1	3	0.481132	0.244292	1	0	0.603053	0	0.370968	0	0.0	1	1
1	0.166667	1	2	0.339623	0.283105	0	1	0.885496	0	0.564516	0	0.0	2	1
2	0.250000	0	1	0.339623	0.178082	0	0	0.770992	0	0.225806	2	0.0	2	1
3	0.562500	1	1	0.245283	0.251142	0	1	0.816794	0	0.129032	2	0.0	2	1
4	0.583333	0	0	0.245283	0.520548	0	1	0.702290	1	0.096774	2	0.0	2	1

# 使用 sklearn 库进行归一化处理
from sklearn.preprocessing import MinMaxScaler

# 重新读取数据
import pandas as pd
data=pd.read_csv(r"heart.csv")

# 连续特征
list_continuous=["age","trestbps","chol","thalach","oldpeak","ca"]

# 归一化
for i in list_continuous:
    min_max_scaler=MinMaxScaler()
    data[i]=min_max_scaler.fit_transform(data[[i]])

data.head()

	age	sex	cp	trestbps	chol	fbs	restecg	thalach	exang	oldpeak	slope	ca	thal	target
0	0.708333	1	3	0.481132	0.244292	1	0	0.603053	0	0.370968	0	0.0	1	1
1	0.166667	1	2	0.339623	0.283105	0	1	0.885496	0	0.564516	0	0.0	2	1
2	0.250000	0	1	0.339623	0.178082	0	0	0.770992	0	0.225806	2	0.0	2	1
3	0.562500	1	1	0.245283	0.251142	0	1	0.816794	0	0.129032	2	0.0	2	1
4	0.583333	0	0	0.245283	0.520548	0	1	0.702290	1	0.096774	2	0.0	2	1

# 标准化处理
from sklearn.preprocessing import StandardScaler

# 重新读取数据
import pandas as pd
data=pd.read_csv(r"heart.csv")

# 连续特征
list_continuous=["age","trestbps","chol","thalach","oldpeak","ca"]

# 标准化
for i in list_continuous:
    scaler=StandardScaler()
    data[i]=scaler.fit_transform(data[[i]])

data.head()

	age	sex	cp	trestbps	chol	fbs	restecg	thalach	exang	oldpeak	slope	ca	thal	target
0	0.952197	1	3	0.763956	-0.256334	1	0	0.015443	0	1.087338	0	-0.714429	1	1
1	-1.915313	1	2	-0.092738	0.072199	0	1	1.633471	0	2.122573	0	-0.714429	2	1
2	-1.474158	0	1	-0.092738	-0.816773	0	0	0.977514	0	0.310912	2	-0.714429	2	1
3	0.180175	1	1	-0.663867	-0.198357	0	1	1.239897	0	-0.206705	2	-0.714429	2	1
4	0.290464	0	0	-0.663867	2.082050	0	1	0.583939	1	-0.379244	2	-0.714429	2	1

@浙大疏锦行