def fetch_magic_candidators(engine, t_day):
# 中证价值回报量化策略指数的样本空间由满足以下条件的沪深 A 股构成：
# （1）非 ST、*ST 股票，非暂停上市股票； （2）非金融类股票。
all_stocks = list(jq.get_all_securities(types=['stock'], date=t_day).index)
#all_stocks = all_stocks[:100]
# 排除金融类的股票
banks = jq.get_industry_stocks('J66', date=t_day)
brokers = jq.get_industry_stocks('J67', date=t_day)
insurances = jq.get_industry_stocks('J68', date=t_day)
others = jq.get_industry_stocks('J69', date=t_day)
exclude = banks + brokers + insurances + others
filtered_1 = []
for code in all_stocks:
if (code in exclude):
print " ... %s 是金融类, 排除..." % code
continue
filtered_1.append(code)
# 排除 ST
st = jq.get_extras('is_st',
filtered_1,
start_date=t_day,
end_date=t_day,
df=False)
filtered_2 = []
for code in filtered_1:
if st[code][0]:
print " ... %s 是ST，排除" % code
continue
filtered_2.append(code)
filtered_3 = []
# 排除停牌
for code in filtered_2:
if data_fetcher.is_paused(engine, code, t_day):
print " ... %s 停牌，排除" % code
continue
filtered_3.append(code)
return filtered_3
def predict(train_cfg):
with open(train_cfg,"r") as f:
train_cfg = json.load(f)
feature_cfg = train_cfg["feature_cfg"]
model_name = train_cfg["model"]
model_cfg = train_cfg["model_cfg"]
#feature_cfg = "./config/feature_create_cfg.json"
with open(feature_cfg,"r") as f:
feature_cfg = json.load(f)
stock_list = jq.get_industry_stocks('I64')
feature_creator = Feature(feature_cfg,stock_list)
#label_creator = Label(feature_cfg,stock_list)
model = restore_model_from_file(model_name,model_cfg)
if not buffered_feature:
feature = []
label = []
for date in feature_creator.date_list:
print ("creating feature, date:",date)
feature_per_day = feature_creator.creatFeatureByDate(date)
feature.append(feature_per_day)
feature = np.array(feature).reshape(-1,feature[0].shape[-1])
print (feature.shape)
np.save("./DATA/buffer/feature_for_prediction.npy",feature)
else:
feature = np.load("./DATA/buffer/feature_for_prediction.npy")
y = model.predict(feature)
print (y.shape)
def record(self, entity, start, end, size, timestamps):
try:
industry_stocks = get_industry_stocks(entity.code, date=now_pd_timestamp())
except:
industry_stocks = get_concept_stocks(entity.code, date=now_pd_timestamp())
if len(industry_stocks) == 0:
return None
df = pd.DataFrame({"stock": industry_stocks})
df["stock_id"] = df.stock.apply(lambda x: to_entity_id(x, "stock"))
df["stock_code"] = df.stock_id.str.split("_", expand=True)[2]
df["stock_name"] = df.stock_id.apply(lambda x:get_data(data_schema=Stock, entity_id=x, provider='joinquant').name)
df["block_type"] = entity.block_type
df["code"] = entity.code
df["name"] = entity.name
df["exchange"] = entity.exchange
df["timestamp"] = now_pd_timestamp()
df["entity_id"] = entity.id
df["entity_type"] = "block"
df["id"] = df.apply(lambda x: x.entity_id + "_" + x.stock_id, axis=1)
if df.empty:
return None
df_to_db(data_schema=self.data_schema, df=df, provider=self.provider,
force_update=True)
self.logger.info('finish recording BlockStock:{},{}'.format(entity.category, entity.name))
def fetch_fundamentals_1_year_300(engine, the_year):
#
# 沪深300 样本股调整实施时间分别是每年 6 月和 12 月的第二个星期五的下一交易日
# 我们假设每年五月(年报出来)换仓
print "fetch fetch_fundamentals for %d" % the_year
yyyymmdd = "%d-05-01" % the_year
# 取当年沪深300成份
hs300 = jq.get_index_stocks('000300.XSHG', date=yyyymmdd)
#print hs300
#J66 货币金融服务 1991-04-03
#J67 资本市场服务 1994-01-10
#J68 保险业 2007-01-09
#J69 其他金融业
# 排除不适用魔法公式的行业
banks = jq.get_industry_stocks('J66', date=yyyymmdd)
brokers = jq.get_industry_stocks('J67', date=yyyymmdd)
insurances = jq.get_industry_stocks('J68', date=yyyymmdd)
others = jq.get_industry_stocks('J69', date=yyyymmdd)
exclude = banks + brokers + insurances + others
#print banks
#print insurances
for code in hs300:
if (code in exclude):
print " ... skip %s ..." % code
continue
print " fetching %s" % code
fetch_target_stock_fundamentals(engine, code, the_year)
#break
print "finished fetching fetch_fundamentals for %d" % the_year
print
def get_industry_stocks():
df = jq.get_industries(name='zjw')
s2i = {}
row_num = len(df.index)
for i in range(row_num):
industry_code = df.index[i]
industry_name = df.iloc[i]['name']
#print industry_code, industry_name
stocks = jq.get_industry_stocks(industry_code)
for stock_code in stocks:
if stock_code in s2i:
s2i[stock_code].append(industry_name)
else:
s2i[stock_code] = [industry_name]
return s2i
daily_label_info = {"info":label,"valid_index":valid_index}
return daily_label_info
def creatDailyLabel(self,date,stock_list):
daily_label_dict = dict()
daily_label_dict["label_all"] = dict()
daily_label_dict["label_all"][date] = self.creatLabelByDate(date,stock_list)
daily_label_dict["date_index"] = {0:date}
return daily_label_dict
def checkLabel():
for creator in self.label_creator_list:
creator.check(self.start_date)
if __name__ == "__main__":
import json
from util.jq_init import login
login()
from data_interface.data_api import UserDataApi
UserDataApi = UserDataApi()
feature_cfg = "./config/feature_create_cfg.json"
with open(feature_cfg,"r") as f:
feature_cfg = json.load(f)
stock_list = jq.get_industry_stocks('I64')
f = Label(feature_cfg,UserDataApi)
label_all = f.createLabelAll(stock_list)
print (label_all["label_all"]["2020-03-04"]["valid_index"])
3.然后观察2021年的市盈率表现
"""
#设置benchmark_date
benchmark_date = '2020-12-31'
# 设置股票范围，000018.XSHG 上证180金融股，XSHG 上海证券交易所
#pool = jq.get_index_stocks('000018.XSHG')
# 获取金融行业板块股票
"""
pool1 = jq.get_industry_stocks('J69', date=benchmark_date)
pool2 = jq.get_industry_stocks('J66', date=benchmark_date)
# 得到所有金融业的股票
pool = pool1 + pool2
"""
# 获取计算机行业板块股票
pool = jq.get_industry_stocks('C39', date=benchmark_date)
pool
# In[19]:
q = jq.query(jq.valuation.code, jq.valuation.pe_ratio, jq.valuation.market_cap) .filter(jq.valuation.pe_ratio > 0, jq.valuation.code.in_(pool)) .order_by(jq.valuation.pe_ratio.asc())
q
df = jq.get_fundamentals(q)
df
# In[20]:
# encoding=utf-8
"""
本脚本用来分析一个行业中龙头股与非龙头股的走势区别
"""
# from JQData_Test.JQ_Industry_Analysis_Sub import plt,jqdatasdk,np
# from JQData_Test.auth_info import *
import jqdatasdk
import numpy as np
from pylab import *
# 下载医药板块的所有stk
stk_list_yiyao = jqdatasdk.get_industry_stocks(industry_code='801150')
# 下载该板块所有stk的close数据
df_yiyao = jqdatasdk.get_price(stk_list_yiyao,
start_date='2010-01-01',
end_date='2018-10-01',
frequency='daily')
df_yiyao_close = df_yiyao['close']
# 龙头股与非龙头股
longtou = ['601607.XSHG', '000623.XSHE']
nolongtou = [x for x in stk_list_yiyao if x not in longtou]
# 非龙头每支stk起始投资1000块
for stk in nolongtou:
df_yiyao_close[stk + '_std'] = df_yiyao_close.apply(
lambda x: x[stk] * (1000 / df_yiyao_close.head(1)[stk].values[0]),
axis=1)
def get_industry_stocks(self, industry_id: str):
return jq.get_industry_stocks(industry_id)
def fetch_brk_candidators(engine,t_day):
brokers = jq.get_industry_stocks( 'J67' , date= t_day )
return brokers