def disconnect(self):
if self.connection:
if self.cursor:
self.cursor.close()
self.connection.close()
logger.info("MYSQL connection is closed")
self.connection = None
self.cursor = None
def test_http_methods(self):
logger.info("Testing Http methods like POST, PUT, DELETE")
result = self.simulate_post('/api/v1/embl/search?name=tbx')
self.assertEqual(result.status, falcon.HTTP_405)
result = self.simulate_put('/api/v1/embl/search?name=tbx')
self.assertEqual(result.status, falcon.HTTP_405)
result = self.simulate_delete('/api/v1/embl/search?name=tbx')
self.assertEqual(result.status, falcon.HTTP_405)
def test_validate_params(self):
logger.info("Validating parameters")
output = {
u"error":
u"Search param 'name' cannot be none or less than 3 chars"
}
result = self.simulate_get('/api/v1/embl/search?name=tb')
self.assertEqual(result.json, output)
def send_mail(email_id, attach_file, to_email, book_title, book_author):
if gk7.SEND_MAIL_TYPE == 'api':
data = {
'email_id': email_id,
'file_path': attach_file,
'to_user': to_email,
'title': book_title,
'author': book_author
}
ApiTask.post.delay(gk7.API_URL.get('mail'), urllib.urlencode(data))
logger.info(u'调用API发送邮件接口中...')
return True
# 其他类型
# 发送邮件并修改待发送邮件状态
MailTask.send.delay(email_id, attach_file, to_email, book_title, book_author)
logger.info(u'发送邮件中...')
return True
def get_data(self, query, params, retry_counter=0) -> []:
try:
logger.info("running query " + query)
self.cursor.execute(query, params)
retry_counter = 0
except DBError as error:
if retry_counter >= Constants.limit_retries.value:
raise error
else:
retry_counter += 1
logger.info("DB Connection error {}. retrying {}".format(
str(error).strip(), retry_counter))
time.sleep(2)
self.reset()
self.get_data(query, retry_counter)
columns = tuple([col[0] for col in self.cursor.description])
data = [dict(zip(columns, row)) for row in self.cursor]
return data
def on_get(self, req, resp):
gene_name = req.get_param(Constants.name.value, None)
species = req.get_param(Constants.species.value, None)
http_status = falcon.HTTP_400
if gene_name and len(gene_name) >= Constants.min_keyword_count.value:
query = """select stable_id as id, display_label as name, species from gene_autocomplete where display_label like %s"""
params = (gene_name.lower()+"%",)
if species:
query+=" and species = %s"
params=(gene_name.lower()+"%",species,)
try:
http_status = falcon.HTTP_200
response_body = json.dumps(self.database.get_data(query,params), ensure_ascii=False)
except Exception as exp:
logger.info(str(exp))
http_status = falcon.HTTP_400
response_body = json.dumps({Constants.error.value: str(exp)})
else:
response_body = json.dumps({Constants.error.value: "Search param 'name' cannot be none or less than 3 chars"})
resp.content_type = falcon.MEDIA_JSON
resp.status = http_status
resp.body = response_body
def connect(self, retry_counter=0):
if not self.connection:
try:
self.connection = connector.connect(
host=Constants.host.value,
user=Constants.user.value,
database=Constants.database.value,
pool_size=Constants.pool_size.value,
connect_timeout=3)
retry_counter = 0
return self.connection
except DBError as error:
if not self.reconnect or retry_counter >= Constants.limit_retries.value:
raise error
else:
retry_counter += 1
logger.info(
"DB Connection error {}. reconnecting {}".format(
str(error).strip(), retry_counter))
time.sleep(5)
self.connect(retry_counter)
except (Exception, DBError) as error:
raise error
def labeling(self, close_df, num_days=5, save=False, min_pct=0.05):
"""labeling func
ラベリングの処理を実行する
Args:
close_df(pandas.DataFrame): 価格情報
num_days(int): xxx
save(bool): 保存するかどうか
min_pct(float): drop_labels で使用
Note:
1. 変化分が大きいシグナルを検出(t_events, get_t_events)、CUSUMサンプリング
2. 変化率を計算(daily_vol_df, get_daily_vol)
3. トリプルバリア法により、ラベルづけ可能なシグナルを検出(events, get_events)
4. ラベルを決定(get_bin)
5. 重複ラベルを削除(drop_labels)
"""
logger.info("close_df len: {}".format(len(close_df)))
t_events = get_t_events(close_df, h=30)
daily_vol_df = self.get_daily_vol(close_df, num_days=num_days)
logger.info("t_events len: {}".format(len(t_events)))
events = self.get_events(close = close_df, t_events=t_events, ptsl=[0.1, 0.1], \
trgt=daily_vol_df, min_ret=0.03, num_threads=1, num_days=num_days,\
t1=True)
logger.info("events len: {}".format(len(events)))
bins = self.get_bins(events, close_df)
bins = self.drop_labels(bins, min_pct=min_pct)
if save:
print("label file を保存します。")
bins.to_csv(self.bins_save_filename)
events.to_csv(self.events_save_filename)
self.close_df = close_df
self.bins = bins
self.events = events
return bins, events
# 设置系统编码
reload(sys)
sys.setdefaultencoding('utf-8')
urls = (
'/','Index',
'/send', 'Send',
'/admin/login', 'Login',
)
#web.config.debug = False
app = web.application(urls, globals())
class Index:
def GET(self):
return json.dumps({'msg':'not services'})
def POST(self):
return json.dumps({'msg':'not servicindex'})
if __name__ == "__main__":
logger.info(u'-----------系统启动-----------')
# 初始化书籍封面目录
if not os.path.exists(gk7.BOOK_COVER_DIRS):
os.makedirs(gk7.BOOK_COVER_DIRS)
#web.wsgi.runwsgi = lambda func, addr=None: web.wsgi.runfcgi(func, addr) ##这行是新增的
app.run()
def sample_weighting(self, check_mc=False, save=False):
"""sample_weighting func
標本の重み付けの処理
Args:
check_mc(bool): モンテカルロ法で逐次ブートストラップ法の有用性を確認するかどうか
"""
# スニペット 4.1
# スニペット 4.2 ラベルの平均性の推定
num_co_events = mp_pandas_obj(func=self.mp_num_co_events,
pd_obj=("molecule", self.events.index),
num_threads=self.num_threads,
close_idx=self.labeling.close_df.index,
t1=self.events["t1"])
num_co_events = num_co_events.loc[~num_co_events.index.duplicated(
keep="last")]
num_co_events = num_co_events.reindex(
self.labeling.close_df.index).fillna(0)
self.bins["tw"] = mp_pandas_obj(self.mp_sample_tw, ("molecule", self.events.index), num_threads=self.num_threads, \
t1=self.events["t1"], num_co_events=num_co_events)
# スニペット 4.10 絶対リターンの帰属による標本ウェイトの決定
self.bins["w"] = mp_pandas_obj(self.mp_sample_w, ("molecule", self.events.index), num_threads=self.num_threads, \
t1=self.events["t1"], num_co_events=num_co_events, close = self.labeling.close_df)
self.bins["w"] *= self.bins.shape[0] / self.bins["w"].sum()
# スニペット 4.11 時間減衰ファクターの実装
self.bins["tw"] = get_time_decay(self.bins["tw"], c_if_last_w=0.5)
# bar_ix = self.get_bar_ix()
bar_ix = self.labeling.close_df.index
if check_mc:
# スニペット 4.8 標準的ブートストラップと逐次ブートストラップによる独自性
num_iters = 2
jobs = []
for i in range(int(num_iters)):
job = {
"func": SampleWeighting.aux_mc,
"t1": self.events["t1"],
"bar_ix": bar_ix
}
jobs.append(job)
if self.num_threads == 1:
out = process_jobs_single(jobs)
out_df = pd.DataFrame(out)
print("out_df", out_df)
print("out_df describe", out_df.describe())
# スニペット 4.3
# スニペット 4.4
ind_matrix = SampleWeighting.get_ind_matrix(bar_ix, self.events["t1"])
phi = SampleWeighting.seq_bootstrap(ind_matrix)
phi = list(set(sorted(phi)))
self.bins_sampled = self.bins.iloc[phi, :]
logger.info("サンプル前 -> サンプル後: {} -> {}".format(len(self.bins),
len(self.bins_sampled)))
if save:
print("結果を保存します")
self.bins_sampled.to_csv(self.bins_save_filename)
return self.bins_sampled
def process_request(self, req, resp):
if req.method not in ["GET"]:
logger.info("")
resp.status = falcon.HTTP_405
resp.body = json.dumps(
{Constants.error.value: "API supports only [GET,]"})
# 设置系统编码
reload(sys)
sys.setdefaultencoding('utf-8')
urls = (
'/','Index',
'/send', 'Send',
'/admin/login', 'Login',
)
#web.config.debug = False
app = web.application(urls, globals())
class Index:
def GET(self):
return json.dumps({'msg':'not services'})
def POST(self):
return json.dumps({'msg':'not servicindex'})
if __name__ == "__main__":
logger.info(u'-----------系统启动-----------')
#web.wsgi.runwsgi = lambda func, addr=None: web.wsgi.runfcgi(func, addr) ##这行是新增的
app.run()
#application = app.wsgifunc()
def test_get_all_records(self):
logger.info("Searching for gene name tbx")
result = self.simulate_get('/api/v1/embl/search?name=tbx')
self.assertEqual(result.status, falcon.HTTP_200)