Traceback (most recent call last)解决方法

response.text 不是标准的json字符串 比如开头缺少{之类的情况

得专门下载一个软件。

import csv

from sklearn.feature_extraction import DictVectorizer
from sklearn import preprocessing
from sklearn import tree
# 加载存储在磁盘上的 决策树
def grabTree(filename):
    import pickle
    fr = open(filename, 'rb')
    return pickle.load(fr)
film_data = open('film.csv', 'rt')

reader = csv.reader(film_data)

# 表头信息
headers = next(reader)
print(headers)

feature_list = [] # 特征值
result_list = []  # 结果

for row in reader:
    result_list.append(row[-1])

    # 去掉首位两列，特征集中只保留'type', 'country', 'gross'
    feature_list.append(dict(zip(headers[1:-1], row[1:-1])))

print(result_list)
print(feature_list)

vec = DictVectorizer()  # 将dict类型的list数据，转换成numpy array
dummyX = vec.fit_transform(feature_list).toarray()
dummyY = preprocessing.LabelBinarizer().fit_transform(result_list)

# country   |,gross|,type
# 0，0，0 , 0 |0，0，|0 , 0，0

# 注意，dummyX是按首字母排序的 'country','gross','type'
print(dummyX)
print(dummyY)

clf = tree.DecisionTreeClassifier(criterion='entropy', random_state=0)
clf = clf.fit(dummyX, dummyY)
print("clf: " + str(clf))

import pydotplus

dot_data = tree.export_graphviz(clf,
                                feature_names=vec.get_feature_names(),
                                filled=True, rounded=True,
                                special_characters=True,
                                out_file=None)
graph = pydotplus.graph_from_dot_data(dot_data)
graph.write_pdf("film.txt")

# # 开始预测
A = ([[0, 0, 0, 1, 0, 1, 0, 1, 0]]) # 日本(4)-低票房(2)-动画片(3)
B = ([[0, 0, 1, 0, 0, 1, 0, 1, 0]]) # 法国(4)-低票房(2)-动画片(3)
C = ([[1, 0, 0, 0, 1, 0, 1, 0, 0]]) # 美国(4)-高票房(2)-动作片(3)
#
predict_result = clf.predict(B)
print("预测结果: " + str(predict_result))