深入浅出python机器学习_6.3.1_随机森林实例——要不要和相亲对象进一步发展

# 6.3.1 数据集的准备

import pandas as pd

data=pd.read_csv('adult.csv',header=None,index_col=False,
                names=['年龄','单位性质','权重','学历','受教育时长',
                       '婚姻状况','职业','家庭状况','种族','性别',
                      '资产所得','资产损失','周工作时长','原籍',
                      '收入'])
data_lite=data[['年龄','单位性质','学历','性别','周工作时长','职业','收入']]

display(data_lite.head())

	年龄	单位性质	学历	性别	周工作时长	职业	收入
0	39	State-gov	Bachelors	Male	40	Adm-clerical	<=50K
1	50	Self-emp-not-inc	Bachelors	Male	13	Exec-managerial	<=50K
2	38	Private	HS-grad	Male	40	Handlers-cleaners	<=50K
3	53	Private	11th	Male	40	Handlers-cleaners	<=50K
4	28	Private	Bachelors	Female	40	Prof-specialty	<=50K

# 6.3.2 用get_dummies处理数据

data_dummies=pd.get_dummies(data_lite)

print('样本原始特征：\n',list(data_lite.columns),'\n')

print('虚拟变量特征：\n',list(data_dummies.columns),'\n')

print('data_dummies.shape:\n',data_dummies.shape,'\n')

print('data_dummies的类型：\n',type(data_dummies))

样本原始特征：
 ['年龄', '单位性质', '学历', '性别', '周工作时长', '职业', '收入'] 

虚拟变量特征：
 ['年龄', '周工作时长', '单位性质_ ?', '单位性质_ Federal-gov', '单位性质_ Local-gov', '单位性质_ Never-worked', '单位性质_ Private', '单位性质_ Self-emp-inc', '单位性质_ Self-emp-not-inc', '单位性质_ State-gov', '单位性质_ Without-pay', '学历_ 10th', '学历_ 11th', '学历_ 12th', '学历_ 1st-4th', '学历_ 5th-6th', '学历_ 7th-8th', '学历_ 9th', '学历_ Assoc-acdm', '学历_ Assoc-voc', '学历_ Bachelors', '学历_ Doctorate', '学历_ HS-grad', '学历_ Masters', '学历_ Preschool', '学历_ Prof-school', '学历_ Some-college', '性别_ Female', '性别_ Male', '职业_ ?', '职业_ Adm-clerical', '职业_ Armed-Forces', '职业_ Craft-repair', '职业_ Exec-managerial', '职业_ Farming-fishing', '职业_ Handlers-cleaners', '职业_ Machine-op-inspct', '职业_ Other-service', '职业_ Priv-house-serv', '职业_ Prof-specialty', '职业_ Protective-serv', '职业_ Sales', '职业_ Tech-support', '职业_ Transport-moving', '收入_ <=50K', '收入_ >50K'] 

data_dummies.shape:
 (32561, 46) 

data_dummies的类型：
 

# 显示数据集中的前五行

#显示所有列
pd.set_option('display.max_columns', None)
#显示所有行
pd.set_option('display.max_rows', None)

data_dummies.head()

	年龄	周工作时长	单位性质_ ?	单位性质_ Federal-gov	单位性质_ Local-gov	单位性质_ Never-worked	单位性质_ Private	单位性质_ Self-emp-inc	单位性质_ Self-emp-not-inc	单位性质_ State-gov	单位性质_ Without-pay	学历_ 10th	学历_ 11th	学历_ 12th	学历_ 1st-4th	学历_ 5th-6th	学历_ 7th-8th	学历_ 9th	学历_ Assoc-acdm	学历_ Assoc-voc	学历_ Bachelors	学历_ Doctorate	学历_ HS-grad	学历_ Masters	学历_ Preschool	学历_ Prof-school	学历_ Some-college	性别_ Female	性别_ Male	职业_ ?	职业_ Adm-clerical	职业_ Armed-Forces	职业_ Craft-repair	职业_ Exec-managerial	职业_ Farming-fishing	职业_ Handlers-cleaners	职业_ Machine-op-inspct	职业_ Other-service	职业_ Priv-house-serv	职业_ Prof-specialty	职业_ Protective-serv	职业_ Sales	职业_ Tech-support	职业_ Transport-moving	收入_ <=50K	收入_ >50K
0	39	40	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
1	50	13	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	1	0
2	38	40	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0
3	53	40	0	0	0	0	1	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	1	0
4	28	40	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	1	0

features=data_dummies.loc[:,'年龄':'职业_ Transport-moving']

print('features的类型：\n',type(features),'\n')

X=features.values

print('X的类型：\n',type(X),'\n')

print('打印X的前五行:\n',X[:5,:])

y=data_dummies['收入_ >50K'].values

print('特征形态：{} 标签形态{}'.format(X.shape,y.shape))


# 避免显示不全：
# import numpy as np
# np.set_printoptions(threshold=np.inf)

features的类型：
  

X的类型：
  

打印X的前五行:
 [[39 40  0  0  0  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0  1  0  0  0
   0  0  0  0  1  0  1  0  0  0  0  0  0  0  0  0  0  0  0  0]
 [50 13  0  0  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0  0  1  0  0  0
   0  0  0  0  1  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0]
 [38 40  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  1  0
   0  0  0  0  1  0  0  0  0  0  0  1  0  0  0  0  0  0  0  0]
 [53 40  0  0  0  0  1  0  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0  0
   0  0  0  0  1  0  0  0  0  0  0  1  0  0  0  0  0  0  0  0]
 [28 40  0  0  0  0  1  0  0  0  0  0  0  0  0  0  0  0  0  0  1  0  0  0
   0  0  0  1  0  0  0  0  0  0  0  0  0  0  0  1  0  0  0  0]]
特征形态：(32561, 44) 标签形态(32561,)

features.columns

type(features.columns)

pandas.core.indexes.base.Index

# 用决策树建模并作出预测
from sklearn.model_selection import train_test_split

X_train,X_test,y_train,y_test=train_test_split(X,y,random_state=0)

from sklearn import tree

go_dating_tree=tree.DecisionTreeClassifier(max_depth=5)

go_dating_tree.fit(X_train,y_train)

print('模型的分：{:.2f}'.format(go_dating_tree.score(X_test,y_test)))

模型的分：0.80