def main(_):
mode = tf.flags.FLAGS.mode
#print("is gpu avaliable:",tf.test.is_gpu_available())
if mode == 'train':
model.train()
elif mode == 'test':
model.test()
elif mode == 'predict':
model.predict()
else:
raise ValueError('--mode {} was not found.'.format(mode))
def trim(model, xtrain, ytrain, name, threshold, path_to_data, win_len):
"""
removes from xtrain, ytrain elements on which the model has F1 greater than threshold
:param path_to_data: path to the folder where the trimmed dataset will be saved
:return: trimmed dataset
"""
pred_train = np.array(model.predict(xtrain))
xtrain_new = xtrain.copy()
ytrain_new = ytrain.copy()
counter = 0
for i in range(len(xtrain)):
pred = pred_train[i, win_len // 2:5000 - win_len // 2, :]
y = ytrain[i, win_len // 2:5000 - win_len // 2, :]
stat = statistics(np.expand_dims(y, axis=0),
np.expand_dims(pred, axis=0))
F = F_score(stat)
if F >= threshold:
xtrain_new = np.delete(xtrain_new, i - counter, axis=0)
ytrain_new = np.delete(ytrain_new, i - counter, axis=0)
counter += 1
if not os.path.exists(path_to_data):
os.makedirs(path_to_data)
outfile = open(path_to_data + "\\trim_" + name + ".pkl", 'wb')
pkl.dump({"x": xtrain_new, "y": ytrain_new}, outfile)
outfile.close()
return xtrain_new, ytrain_new
async def predict(payload: FakeNewsPayload):
vector = fake_news_preprocess(payload)
prediction = model.predict(vector)
real_probability = float(prediction[0][0])
fake_probability = 1 - real_probability
response = {'fake_probability': fake_probability}
return response
def predict_cost(model, dicti):
di = {
'AGE': 0,
'LAT': 0,
'LON': 0,
'ethnicity_hispanic': 0.0,
'ethnicity_nonhispanic': 0.0,
'gender_F': 0.0,
'gender_M': 0.0,
'marital_M': 0.0,
'marital_S': 0.0,
'race_asian': 0.0,
'race_black': 0.0,
'race_native': 0.0,
'race_white': 0.0,
'reasoncode_10509002.0': 0.0,
'reasoncode_185086009.0': 0.0,
'reasoncode_192127007.0': 0.0,
'reasoncode_195662009.0': 0.0,
'reasoncode_201834006.0': 0.0,
'reasoncode_230265002.0': 0.0,
'reasoncode_233678006.0': 0.0,
'reasoncode_239872002.0': 0.0,
'reasoncode_239873007.0': 0.0,
'reasoncode_24079001.0': 0.0,
'reasoncode_254837009.0': 0.0,
'reasoncode_26929004.0': 0.0,
'reasoncode_301011002.0': 0.0,
'reasoncode_363406005.0': 0.0,
'reasoncode_36971009.0': 0.0,
'reasoncode_38822007.0': 0.0,
'reasoncode_40275004.0': 0.0,
'reasoncode_424132000.0': 0.0,
'reasoncode_43878008.0': 0.0,
'reasoncode_44054006.0': 0.0,
'reasoncode_444814009.0': 0.0,
'reasoncode_55822004.0': 0.0,
'reasoncode_59621000.0': 0.0,
'reasoncode_75498004.0': 0.0,
'reasoncode_87433001.0': 0.0,
'reasoncode_88805009.0': 0.0,
'reasoncode_90560007.0': 0.0
}
# df=pd.DataFrame(di)
di['AGE'] = (dicti['AGE'])
di['LAT'] = (dicti['LAT'])
di['LON'] = (dicti['LON'])
di['ethnicity_' + dicti['ethnicity']] = 1.0
di['gender_' + dicti['gender']] = 1.0
di['marital_' + dicti['marital']] = 1.0
di['race_' + dicti['race']] = 1.0
di['reasoncode_' + dicti['reasoncode'] + '.0'] = 1.0
return model.predict(pd.DataFrame(di, index=[0]))[0]
def predict():
# get data
data = request.get_json(force=True)
# convert data into dataframe
data.update((x, [y]) for x, y in data.items())
data_df = pd.DataFrame.from_dict(data)
# predictions
result = model.predict(data_df)
# send back to browser
output = {'results': float(result)}
# return data
return jsonify(results=output)
def trim(model, xtrain, ytrain, data_name, threshold, path_to_data, win_len):
pred_train = np.array(model.predict(xtrain))
xtrain_new = xtrain.copy()
ytrain_new = ytrain.copy()
counter = 0
for i in range(len(xtrain)):
pred = pred_train[i, win_len // 2:5000 - win_len // 2, :]
y = ytrain[i, win_len // 2:5000 - win_len // 2, :]
stat = statistics(np.expand_dims(y, axis=0),
np.expand_dims(pred, axis=0))
F = F_score(stat)
if F >= threshold:
xtrain_new = np.delete(xtrain_new, i - counter, axis=0)
ytrain_new = np.delete(ytrain_new, i - counter, axis=0)
counter += 1
outfile = open(path_to_data + "\\trim_" + data_name + ".pkl", 'wb')
pkl.dump({"x": xtrain_new, "y": ytrain_new}, outfile)
outfile.close()
return xtrain_new, ytrain_new
def simulate(test_np_x, test_np_y):
"""
模拟购买彩票，对测试数据进行回测
:param test_np_x: 测试数据输入
:param test_np_y: 测试数据输出
:return: 本次模拟的净收益
"""
# 获得的奖金总额
money_in = 0
# 买彩票花出去的钱总额
money_out = 0
# 预测
predicts = model.predict(test_np_x, batch_size=settings.BATCH_SIZE)
# 共有多少组数据
samples_num = len(test_np_x['x1'])
# 对于每一组数据
for j in range(samples_num):
# 这一期的真实开奖结果
outputs = []
for k in range(settings.FRONT_SIZE + settings.BACK_SIZE):
outputs.append(np.argmax(test_np_y['y{}'.format(k + 1)][j]))
# 每一期彩票买五注
money_out += 10
for k in range(5):
# 存放每个球的概率分布的list
probabilities = []
# 对于每一种球，将其概率分布加入到列表中去
for i in range(settings.FRONT_SIZE + settings.BACK_SIZE):
probabilities.append(predicts[i][j])
# 根据概率分布随机选择一个序列
balls = utils.select_seqs(probabilities)
# 计算奖金
award = utils.lotto_calculate(outputs, balls)
money_in += award
if award:
print('{} 中奖了,{}元！ {}/{}'.format(j, award, money_in, money_out))
print('买彩票花费金钱共{}元，中奖金额共{}元，赚取{}元'.format(money_out, money_in, money_in - money_out))
return money_in - money_out
def predict_test(file_path, model, X_test):
model.load_weights(file_path)
prediction = model.predict(X_test, verbose=1, batch_size=32)
return prediction
# @Author : AaronJny
# @Date : 2019/11/26
# @Desc : 指定一个训练好的模型参数，让模型随机选出下期彩票号码
from dataset import LottoDataSet
from models import model
import settings
import utils
# 加载模型参数
model.load_weights(settings.PREDICT_MODEL_PATH)
# 构建数据集
lotto_dataset = LottoDataSet()
# 提取倒数第MAX_STEPS期到最近一期的数据，作为预测的输入
x = lotto_dataset.predict_data
# 开始预测
predicts = model.predict(x, batch_size=1)
# 存放选号结果的列表
result = []
# 存放每个球的概率分布的list
probabilities = [predict[0] for predict in predicts]
# print(probabilities)
# 总共要选出settings.PREDICT_NUM注彩票
for i in range(settings.PREDICT_NUM):
# 根据概率分布随机选择一个序列
balls = utils.select_seqs(probabilities)
# 加入到选号列表中,注意，我们需要把全部的数字+1，恢复原始的编号
result.append([ball + 1 for ball in balls])
# 输出要买的彩票序列
print('本次预测结果如下：')
for index, balls in enumerate(result, start=1):
print('第{}注 {}'.format(index, ' '.join(map(str, balls))))
# @Desc : 自己输入句子测试模型是否有效
from dataset import tokenizer
from models import model
import settings
# 加载训练好的参数
model.load_weights(settings.BEST_WEIGHTS_PATH)
print('启动验证程序！')
while True:
try:
sentence = input('请输入一句话，模型将判断其情绪倾向：')
token_ids, segment_ids = tokenizer.encode(sentence)
output = model.predict([[
token_ids,
], [
segment_ids,
]])[0][0]
if output > 0.5:
print('正面情绪！')
else:
print('负面情绪！')
except KeyboardInterrupt:
print('结束程序！')
break
"""
请输入一句话，模型将判断其情绪倾向：虽然没有买到想要的东西，但我并不沮丧
正面情绪！
请输入一句话，模型将判断其情绪倾向：没有买到想要的东西， 有点沮丧
负面情绪！
请输入一句话，模型将判断其情绪倾向：书挺好的，就是贵了点
def predict_from_real_images(model, image_dims, images, image_labels):
from models.model import predict
model = model.resize(image_dims)
return predict(model, images, image_labels=image_labels)
def predict_careplan(model, dicti):
di = {
'AGE': 0,
'LAT': 0,
'LON': 0,
'ethnicity_hispanic': 0.0,
'ethnicity_nonhispanic': 0.0,
'gender_F': 0.0,
'gender_M': 0.0,
'marital_M': 0.0,
'marital_S': 0.0,
'race_asian': 0.0,
'race_black': 0.0,
'race_native': 0.0,
'race_other': 0.0,
'race_white': 0.0,
'reasoncode_10509002.0': 0.0,
'reasoncode_109838007.0': 0.0,
'reasoncode_110030002.0': 0.0,
'reasoncode_126906006.0': 0.0,
'reasoncode_15724005.0': 0.0,
'reasoncode_15777000.0': 0.0,
'reasoncode_16114001.0': 0.0,
'reasoncode_185086009.0': 0.0,
'reasoncode_192127007.0': 0.0,
'reasoncode_201834006.0': 0.0,
'reasoncode_230265002.0': 0.0,
'reasoncode_233678006.0': 0.0,
'reasoncode_239720000.0': 0.0,
'reasoncode_239872002.0': 0.0,
'reasoncode_239873007.0': 0.0,
'reasoncode_24079001.0': 0.0,
'reasoncode_262574004.0': 0.0,
'reasoncode_263102004.0': 0.0,
'reasoncode_26929004.0': 0.0,
'reasoncode_283371005.0': 0.0,
'reasoncode_283385000.0': 0.0,
'reasoncode_284549007.0': 0.0,
'reasoncode_284551006.0': 0.0,
'reasoncode_301011002.0': 0.0,
'reasoncode_307731004.0': 0.0,
'reasoncode_30832001.0': 0.0,
'reasoncode_33737001.0': 0.0,
'reasoncode_359817006.0': 0.0,
'reasoncode_363406005.0': 0.0,
'reasoncode_36923009.0': 0.0,
'reasoncode_370143000.0': 0.0,
'reasoncode_370247008.0': 0.0,
'reasoncode_38822007.0': 0.0,
'reasoncode_39848009.0': 0.0,
'reasoncode_40275004.0': 0.0,
'reasoncode_403190006.0': 0.0,
'reasoncode_403191005.0': 0.0,
'reasoncode_424132000.0': 0.0,
'reasoncode_44054006.0': 0.0,
'reasoncode_444448004.0': 0.0,
'reasoncode_444470001.0': 0.0,
'reasoncode_44465007.0': 0.0,
'reasoncode_449868002.0': 0.0,
'reasoncode_45816000.0': 0.0,
'reasoncode_47505003.0': 0.0,
'reasoncode_55680006.0': 0.0,
'reasoncode_55822004.0': 0.0,
'reasoncode_58150001.0': 0.0,
'reasoncode_59621000.0': 0.0,
'reasoncode_62106007.0': 0.0,
'reasoncode_62564004.0': 0.0,
'reasoncode_65966004.0': 0.0,
'reasoncode_67811000119102.0': 0.0,
'reasoncode_69896004.0': 0.0,
'reasoncode_70704007.0': 0.0,
'reasoncode_72892002.0': 0.0,
'reasoncode_87433001.0': 0.0,
'reasoncode_88805009.0': 0.0,
'reasoncode_90560007.0': 0.0,
'reasoncode_93761005.0': 0.0,
'reasoncode_94260004.0': 0.0,
'reasoncode_95417003.0': 0.0
}
di['AGE'] = float(dicti['AGE'])
di['LAT'] = float(dicti['LAT'])
di['LON'] = float(dicti['LON'])
di['ethnicity_' + dicti['ethnicity']] = 1.0
di['gender_' + dicti['gender']] = 1.0
di['marital_' + dicti['marital']] = 1.0
di['race_' + dicti['race']] = 1.0
di['reasoncode_' + dicti['reasoncode'] + '.0'] = 1.0
return model.predict(pd.DataFrame(di, index=[0]))[0]
