python决策树实例教程_Python：决策树分析的一个实例,python,案例

#23 pythoh descison tree

from sklearn.feature_extraction import DictVectorizer

import csv

import pandas as pd

from sklearn import preprocessing

from sklearn import tree

from sklearn.externals.six import StringIO

data1=pd.read_csv('D://allelectronics.csv')

data1

data1.shape[0]

column_list=data1.columns.tolist()

column_list=column_list[0].split(';')

column_list

data2=data1.iloc[:,0].str.split(';',expand=True)

data2.columns=column_list

data2

featrue_list=[]

label_list=[]

label_list=data2['class:buys_computer']

label_list

len(data2)

t=0

while t<= len(data2):

rowdict={}

for i in range(1,len(data2.columns)-1):

print(i)

rowdict[data2.columns[i]]=data2.iloc[t,i]

featrue_list.append(rowdict)

t+=1

featrue_list

vec=DictVectorizer()

dummy_x=vec.fit_transform(featrue_list).toarray()

print('dummyx'+str(dummy_x))

print(vec.get_feature_names())

lb=preprocessing.LabelBinarizer()

dumm_y=lb.fit_transform(label_list)

dumm_y

clf=tree.DecisionTreeClassifier(criterion='entropy')

clf.fit(dummy_x,dumm_y)

clf

# from IPython.display import Image

# dot_data = tree.export_graphviz(clf, out_file=None,

# feature_names=featrue_list,

# class_names=label_list,

# )

# graph = pydotplus.graph_from_dot_data(dot_data)

# Image(graph.create_png())

with open('x1.dot','w') as f:

f=tree.export_graphviz(clf,feature_names=vec.get_feature_names(),out_file=f)

# import pydotplus

# graph=pydotplus.graph_from_dot_data(f)

# graph.write_pdf('1.pdf')

onerow=dummy_x[0,:]

onerow

print('onerowx'+str(onerow))

onerow[0]=1

onerow[1]=0

import numpy as np

onerow=np.array(onerow).reshape((1,-1))

PredictedY=clf.predict(onerow)

PredictedY

关注

回复 ele 获取源数据