def store_classifier(clf, trainset, filepath):
classifier = SklearnClassifier(clf)
classifier.train(trainset)
pred = classifier.prob_classify_many(extract_features(sentiment))
p_file = open(filepath,'w+') #把结果写入文档
# for i in pred:
# p_file.write(str(i.prob('pos'))+' '+str(i.prob('neg')))
for (i,j) in zip(pred,sen_cur):
p_file.write(str(i.prob('pos'))+'\t'+str(i.prob('neg'))+'\t'+j + '\n')
p_file.close()
def plot_ROC(classifier):
cv = StratifiedKFold(n_splits=6) # 导入该模型，后面将数据划分6份
classifier = SklearnClassifier(classifier)
global k
# 画平均ROC曲线的两个参数
mean_tpr = 0.0 # 用来记录画平均ROC曲线的信息
mean_fpr = np.linspace(0, 1, 100)
cnt = 0
for i, (train, test) in enumerate(cv.split(
dev_roc, tag_dev_roc)): # 利用模型划分数据集和目标变量 为一一对应的下标
cnt += 1
devtest1 = np.array(for_roc_plot)
classifier.train(devtest1[train])
# print(cnt)
# print(train)
dev1 = np.array(dev_roc)
tag_dev1 = np.array(tag_dev_roc)
pred_ = classifier.prob_classify_many(dev1[test]) # 测试集的概率
# print(pred_)
# probas_ = classifier.fit(dev1[train], tag_dev1[train]).predict_proba(dev1[test]) # 训练模型后预测每条样本得到两种结果的概率
probas_ = []
for j in pred_:
probas_.append(j.prob('pos'))
fpr, tpr, thresholds = roc_curve(
tag_dev1[test], probas_,
pos_label='pos') # 该函数得到伪正例、真正例、阈值，这里只使用前两个
# print(fpr)
mean_tpr += np.interp(
mean_fpr, fpr,
tpr) # 插值函数 interp(x坐标,每次x增加距离,y坐标) 累计每次循环的总值后面求平均值
mean_tpr[0] = 0.0 # 将第一个真正例=0 以0为起点
roc_auc = auc(fpr, tpr) # 求auc面积
plt.plot(fpr,
tpr,
lw=1,
label='ROC fold {0:.2f} (area = {1:.2f})'.format(
i, roc_auc)) # 画出当前分割数据的ROC曲线
plt.plot([0, 1], [0, 1], '--', color=(0.6, 0.6, 0.6), label='Luck') # 画对角线
mean_tpr /= cnt # 求数组的平均值
mean_tpr[-1] = 1.0 # 坐标最后一个点为（1,1） 以1为终点
mean_auc = auc(mean_fpr, mean_tpr)
plt.plot(mean_fpr,
mean_tpr,
'k--',
label='Mean ROC (area = {0:.2f})'.format(mean_auc),
lw=2)
plt.xlim([-0.05, 1.05]) # 设置x、y轴的上下限，设置宽一点，以免和边缘重合，可以更好的观察图像的整体
plt.ylim([-0.05, 1.05])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate') # 可以使用中文，但需要导入一些库即字体
plt.title(
'Receiver operating characteristic curve using {0}'.format(classifier))
plt.legend(loc="lower right")
fig = plt.gcf()
plt.show()
fig.savefig('../out/ROC_curves/第{0}张'.format(k + 1), dpi=400) #保存图片
k = k + 1
print("Labels :", LogisticRegression_classifier.labels())
print("Type : ", type(LogisticRegression_classifier))
dictum = [tupl[0] for tupl in testing_set]
print("dictum : ", dictum)
try:
print("classify many:",
LogisticRegression_classifier.classify_many(dictum))
except:
print("classify many erreur \n", "Type testing_set: ",
type(dictum), "\n testing_set :", dictum)
try:
print("prob_classify_many:",
LogisticRegression_classifier.prob_classify_many(dictum))
for probdisti in LogisticRegression_classifier.prob_classify_many(
dictum):
list_of_samples = probdisti.samples()
for sample in list_of_samples:
print("Sample: ", sample, " Prob : ",
probdisti.prob(sample))
except:
print("prob_classify_many erreur \n", "Type testing_set:",
type(dictum))
save_classifier = open("LogisticRegression_classifier.pickle",
"wb")
pickle.dump(LogisticRegression_classifier, save_classifier)
save_classifier.close()
'''
