def test_maxent_iteration(self):
print("MaxEntClassifier iteration")
print("---" * 45)
print("Train num = %s" % self.train_num)
print("Test num = %s" % self.test_num)
print("maxiter = %s" % self.max_iter)
from spa.classifiers import MaxEntClassifier
m = MaxEntClassifier(self.max_iter)
iter_results = m.test(self.train_data, self.train_labels,
self.best_words, self.test_data)
filepath = "f_runout/MaxEnt-iteration-%s-train-%d-test-%d-f-%d-maxiter-%d-%s.xls" % \
(self.type,
self.train_num,
self.test_num,
self.feature_num,
self.max_iter,
datetime.datetime.now().strftime(
"%Y-%m-%d-%H-%M-%S"))
results = []
for i in range(len(iter_results)):
try:
results.append(
get_accuracy(self.test_labels, iter_results[i],
self.parameters))
except ZeroDivisionError:
print("ZeroDivisionError")
Write2File.write_contents(filepath, results)
def test_maxent_iteration(self):
print("MaxEntClassifier iteration")
print("---" * 45)
print("Train num = %s" % self.train_num)
print("Test num = %s" % self.test_num)
print("maxiter = %s" % self.max_iter)
from spa.classifiers import MaxEntClassifier
m = MaxEntClassifier(self.max_iter)
iter_results = m.test(self.train_data, self.train_labels, self.best_words, self.test_data)
filepath = "f_runout/MaxEnt-iteration-%s-train-%d-test-%d-f-%d-maxiter-%d-%s.xls" % \
(self.type,
self.train_num,
self.test_num,
self.feature_num,
self.max_iter,
datetime.datetime.now().strftime(
"%Y-%m-%d-%H-%M-%S"))
results = []
for i in range(len(iter_results)):
try:
results.append(get_accuracy(self.test_labels, iter_results[i], self.parameters))
except ZeroDivisionError:
print("ZeroDivisionError")
Write2File.write_contents(filepath, results)
def write(self, filepath, classify_labels, i=-1):
results = get_accuracy(self.test_labels, classify_labels, self.parameters)
if i >= 0:
self.precisions[i][0] = results[10][1] / 100
self.precisions[i][1] = results[7][1] / 100
Write2File.write_contents(filepath, results)
def write(self, filepath, classify_labels, i=-1):
results = get_accuracy(self.test_labels, classify_labels, self.parameters)
if i >= 0:
self.precisions[i][0] = results[10][1] / 100
self.precisions[i][1] = results[7][1] / 100
Write2File.write_contents(filepath, results)
def test_dict():
"""
test the classifier based on Sentiment Dict
"""
print("DictClassifier")
print("---" * 45)
from spa.classifiers import DictClassifier
ds = DictClassifier()
# 对一个单句进行情感分析
# a_sentence = "剁椒鸡蛋好咸,土豆丝很好吃" # result值: 修改前(1)/修改后(1)
# a_sentence = "要是米饭再多点儿就好了" # result值: 修改前(1)/修改后(0)
# a_sentence = "要是米饭再多点儿就更好了" # result值: 修改前(0)/修改后(0)
# a_sentence = "不太好吃，相当难吃，要是米饭再多点儿就好了" # result值: 修改前(1)/修改后(0)
# result = ds.analyse_sentence(a_sentence)
# print(result)
# 对一个文件内语料进行情感分析
corpus_filepath = "D:/My Data/NLP/SA/waimai/positive_corpus_v1.txt"
runout_filepath_ = "f_runout/f_dict-positive_test.txt"
pos_results = ds.analysis_file(corpus_filepath,
runout_filepath_,
start=3000,
end=4000 - 1)
#
# corpus_filepath = "D:/My Data/NLP/SA/waimai/negative_corpus_v1.txt"
# runout_filepath_ = "f_runout/f_dict-negative_test.txt"
# neg_results = ds.analysis_file(corpus_filepath, runout_filepath_, start=3000, end=4000-1)
#
origin_labels = [1] * 1000 + [0] * 1000
classify_labels = pos_results + neg_results
#
# print(len(classify_labels))
#
filepath = "f_runout/Dict-waimai-%s.xls" % (
datetime.datetime.now().strftime("%Y-%m-%d-%H-%M-%S"))
results = get_accuracy(origin_labels, classify_labels, [1000, 1000, 0])
#
Write2File.write_contents(filepath, results)