def re_limit_up(start_date, end_date):
rawdata = read_data('daily', start_date=start_date, end_date=end_date)
# list_days=basic().list_days(rawdata)
limit = read_data('stk_limit', start_date=start_date, end_date=end_date)
data = rawdata.merge(limit[['ts_code', 'trade_date', 'up_limit']],
on=['ts_code', 'trade_date'])
print(data.shape)
data['red_line'] = data.apply(lambda x: 1
if x['low'] == x['up_limit'] else 0,
axis=1)
data['reback_limit'] = data.apply(
lambda x: 1 if (x['open'] == x['close']) &
(x['open'] == x['up_limit']) & (x['low'] < x['close']) else 0,
axis=1)
print(data.shape)
pre = basic().pre_data(data, label=['red_line'], pre_days=2)
data = data.merge(pre[['ts_code', 'trade_date', 'pre_2_red_line']],
on=['ts_code', 'trade_date'])
pre = basic().pre_data(data, label=['reback_limit'], pre_days=1)
data = data.merge(pre[['ts_code', 'trade_date', 'pre_1_reback_limit']],
on=['ts_code', 'trade_date'])
print(data.shape)
data = data.loc[(data['pre_2_red_line'] == 1)
& (data['pre_1_reback_limit'] == 1)]
print(data.shape)
#
save_data('破板后回封表现.csv')
# data=data.loc[(data['up_limit']==data['close'])&(data['open']==data['close'])&(data['low']<data['up_limit'])]
# print(data.shape)
data = data.merge(basic().list_days(data, list_days=15))
print(data.shape)
wool = sheep.wool(data, rawdata)
return wool
def verify(up,data):
t1 = datetime.datetime.now()
s1=sheep.wool(up,data)
t2= datetime.datetime.now()
s2=sheep.wool2(up,data)
t3 = datetime.datetime.now()
s1.to_csv('s1.csv')
s2.to_csv('s2.csv')
print()
def first_up(raw_data,start_date='20200201', end_date='20200215', PRICEB='open', NEWPRICEB='price_buy', PRICES='close',
sell_date=1,
buy_date=2, CUTS=2,CHANGE=['open', 'pre_close'],
model=None, model3_pct=1.01):
if buy_date != 1:
price_buy = basic.basic().pre_data(
raw_data[['ts_code', 'trade_date', PRICEB, '%s/%s' % (CHANGE[0], CHANGE[1])]],
label=[PRICEB, '%s/%s' % (CHANGE[0], CHANGE[1])],
new_label=[NEWPRICEB, 'open_pre_close_change', ],
pre_days=-buy_date + 1)
raw_data = raw_data.merge(price_buy[['ts_code', 'trade_date', NEWPRICEB, 'open_pre_close_change']],
on=['ts_code', 'trade_date'])
# 筛选收盘涨停数据
data = raw_data.loc[:, ['ts_code', 'trade_date', 'close', 'pre_close', 'up_limit', 'open_pre_close_change']]
print(data.shape)
data = data.merge(
basic.basic().pre_data(data[['ts_code', 'trade_date', 'up_limit']], label=['up_limit'],
new_label=['pre_up_limit'])[
['ts_code', 'trade_date', 'pre_up_limit']], on=['ts_code', 'trade_date'])
print(data.shape)
data.dropna(inplace=True)
print(data.shape)
data = data[data['close'] == data['up_limit']]
# print(data.shape)
# raw_data=raw_data[raw_data["ts_code"].isin(data['ts_code'].unique()) == True]
# print('不在涨停df中的股票out',raw_data.shape)
# print(raw_data['ts_code'].unique().shape)
# buy_data = raw_data.loc[:, ['ts_code', 'trade_date', PRICES]]
# z1=sheep.wool(data, raw_data, PRICEB='price_buy')
# print(z1.iloc[-1, -1])
# 全市场，首板，涨停回溯
# z2 = sheep.wool(data, raw_data)
# print(z2.iloc[-1, -1])
# # t1=sheep.wool(data[data['pre_close'] != data['pre_up_limit']],
# # raw_data, PRICEB='price_buy')
# # print(t1.iloc[-1, -1])
# t2 = sheep.wool(data[data['pre_close'] != data['pre_up_limit']],
# raw_data)
# print(t2.iloc[-1, -1])
# data.to_csv('%s~%slimit_stock_list.csv' % (start_date, end_date))
# res_all = pd.DataFrame()
if model:
PRICES = 'price_sell'
if model == 1:
raw_data[PRICES] = raw_data.apply(
lambda x: x['open'] if x['open'] >= x['pre_close'] else x['pre_close'] if x['pre_close'] <= x[
'high'] else x['close'],axis=1)
if model == 2:
raw_data[PRICES] = 10 * raw_data['amount'] / raw_data['vol']
if model==3:
raw_data[PRICES]=raw_data.apply(lambda x:x['open'] if x['open'] >(x['pre_close']*1.01) else x['close'],axis=1)
if model==4:
raw_data[PRICES]=raw_data.apply(lambda x:x['open'] if x['open']>=(1.03*x['pre_close']) else x['open']*1.01 if x['high']>(x['open']*1.01) else x['close'] ,axis=1)
# temp = basic.basic().pre_data(raw_data[['ts_code', 'trade_date', 'ma']], label=['ma'], new_label=[PRICES])
# raw_data = raw_data.merge(temp[['ts_code', 'trade_date', PRICES]],on=['ts_code', 'trade_date'])
res_some = pd.DataFrame()
for cut in range(-12, 11, CUTS):
if cut < -10:
buy_data = data[data['open_pre_close_change'] == -9999]
print('open down')
elif cut > 8:
buy_data = data[data['open_pre_close_change'] == 9999]
print('open up')
else:
buy_data = data[(data['open_pre_close_change'] < cut + 2) & (data['open_pre_close_change'] >= cut)]
print('%s<=x<%s' % (cut, cut + CUTS))
if buy_data.empty:
# res_all.loc['%s~%s' % (cut, cut + CUTS), 'res'] = None
res_some.loc['%s~%s' % (cut, cut + CUTS), 'res'] = None
print(buy_data['open_pre_close_change'].max(), buy_data['open_pre_close_change'].min(),
buy_data['open_pre_close_change'].mean())
continue
# wool_all = sheep.wool(buy_data, raw_data, PRICEB='price_buy', PRICES=PRICES, days=sell_date)
# if wool_all.iloc[-1, -1]:
# wool_all.to_csv('pct%swool_all%s~%s.csv' % (cut, start_date, end_date))
# print('all%s' % cut, wool_all.iloc[-1, -1])
# res_all.loc['%s~%s' % (cut, cut + CUTS), 'res'] = wool_all.iloc[-1, -1]
# res_all.loc['%s~%s' % (cut, cut + CUTS), 'n_mean'] = wool_all['n'].mean()
# res_all.loc['%s~%s' % (cut, cut + CUTS), 'days'] = wool_all.shape[0]
wool_some = sheep.wool(buy_data[buy_data['pre_close'] != buy_data['pre_up_limit']], raw_data,
PRICEB='price_buy', PRICES=PRICES, days=sell_date)
if wool_some.empty:
continue
if wool_some.iloc[-1, -1]:
wool_some.to_csv('pct%swool_some%s~%s-%s.csv' % (cut, start_date, end_date,sell_date))
print('some', wool_some.iloc[-1, -1])
res_some.loc['%s~%s' % (cut, cut + CUTS), 'res'] = wool_some.iloc[-1, -1]
res_some.loc['%s~%s' % (cut, cut + CUTS), 'n_mean'] = wool_some['n'].mean()
res_some.loc['%s~%s' % (cut, cut + CUTS), 'days'] = wool_some.shape[0]
# buy_data.to_csv('pct%slimit_stock_list%s~%s.csv' % (cut, start_date, end_date))
# res_all.to_csv('cut%s-of-all(%s,%s)%s~%s.csv' % (CUTS, PRICEB, PRICES, start_date, end_date))
# res_some.to_csv('cut%s-of-first(%s,%s)%s~%s.csv' % (CUTS, PRICEB, PRICES, start_date, end_date))
return res_some
def idea7(XISHU=0.998, UPPER=False, PERIOD=5, TIMES=5, CHGPCT=[1, 2]):
FORMAT = lambda x: '%.4f' % x
label = ['low_ma5']
path = 'D:\\workgit\\stock\\util\\stockdata\\'
# pct=list(range(-11,11))
today = datetime.datetime.today().date()
tool = basic.basic()
pro = ts.pro_api()
while (not os.path.isfile(path + str(today) + '\data.csv')) or (
not os.path.isfile(path + str(today) + '\daily-basic.csv')):
today = today - datetime.timedelta(1)
# 基础数据，市值信息，
data = pd.read_csv(path + str(today) + '\data.csv',
index_col=0,
dtype={'trade_date': object})[[
'ts_code', 'trade_date', 'close', 'pct_chg', 'low',
'ma5'
]]
# data=data[data['trade_date']>='20191201']
data['lowma5'] = data.apply(lambda x: 1
if x['low'] > (XISHU * x['ma5']) else 0,
axis=1)
if UPPER:
data['lowma5'] = data.apply(lambda x: 1 if (x['lowma5'] == 1) &
(x['low'] <=
(XISHU + UPPER) * x['ma5']) else 0,
axis=1)
stock_baisc = pd.read_csv(
path + str(today) + '\daily-basic.csv',
index_col=0,
dtype={'trade_date':
object})[['ts_code', 'trade_date', 'turnover_rate', 'total_mv']]
history = pd.read_csv(path + str(today) + '\history_name.csv',
index_col=0,
dtype={'trade_date': object})
history['name'] = 'st'
path = path + str(today) + 'idea7\\'
if not os.path.isdir(path):
os.makedirs(path)
print(path)
today = today.strftime('%Y%m%d')
df = data.copy()
df.sort_values(by='trade_date', inplace=True)
# df.to_csv(path+'2019datalowma5.csv')
c_times = df.groupby('ts_code')['lowma5'].rolling(PERIOD).sum()
c_times.index = c_times.index.droplevel()
c_times = pd.DataFrame(c_times)
c_times.rename(columns={'lowma5': 'count'}, inplace=True)
# count_times=count_times[count_times['%s_%s_uptimes'%(label,period)]>=up_period]
df = df.join(c_times)
df = df.merge(history, on=['ts_code', 'trade_date'], how='left')
print(df.shape)
df = df[df['name'].isna()]
df.drop(columns='name', inplace=True)
print(df.shape)
df.dropna(inplace=True)
#
# df.to_csv(path + 'data.csv')
if CHGPCT:
df = df[(df['count'] >= TIMES) & (df['pct_chg'] >= CHGPCT[0]) &
(df['pct_chg'] <= CHGPCT[1])][['ts_code', 'trade_date']]
if df.empty:
return pd.DataFrame()
else:
df.to_csv(path + 'low%sma%stimes%s(%s-%s).csv' %
(XISHU, PERIOD, TIMES, CHGPCT[0], CHGPCT[1]))
wool = sheep.wool(df, data)
if not wool.empty:
wool.to_csv(
path + 'low%s-%sma%stimes%s(%s-%s)huisuxiaoguo.csv' %
(XISHU, XISHU + UPPER, PERIOD, TIMES, CHGPCT[0], CHGPCT[1]))
res = df[df['trade_date'] == today][['ts_code']]
if res.empty:
print('%s无数据' % today)
else:
res.to_csv(path + 'low%sma%stimes%s(%s-%s)%s.txt' %
(XISHU, PERIOD, TIMES, CHGPCT[0], CHGPCT[1], today),
index=False)
return wool
# median_list=filter.StockFilter().stock_basic(name="st|ST|药",market="科创板",industry='生物制药|医药商业|医疗保健|中成药|化学制药')
# print(median_list)
# df[df["ts_code"].isin(median_list['ts_code']) == False][['ts_code']].to_csv('0203非跌停非医药.txt',index=False)
# df[['ts_code']].to_csv('0203非跌停.txt',index=False)
df = df[df['ts_code'].isin(up_data['ts_code']) == False]
print(df.shape)
res = pd.DataFrame()
for days in range(1, NDAY + 1):
for pct in np.arange(-10, 11, t):
data = df[(df['pct_chg'] >= pct) & (df['pct_chg'] < (pct + t))]
# if pct==9:
# print(data)
data_res = sheep.wool(data, raw_data, days=days)
if data_res.empty:
continue
res.loc[pct, 'huisu%s' % days] = data_res.iloc[-1, 0]
# if 'count' not in res.columns:
if data_res.iloc[-1, 1]:
res.loc[pct, 'count'] = data_res.iloc[-1, 1]
down_res = sheep.wool(down_data, raw_data, days=days)
if down_res.empty:
continue
up_res = sheep.wool(up_data, raw_data, days=days)
res.loc['down', 'huisu%s' % days] = down_res.iloc[-1, 0]
res.loc['down', 'count'] = down_res.iloc[-1, 1]
res.loc['up', 'huisu%s' % days] = up_res.iloc[-1, 0]
res.loc['up', 'count'] = up_res.iloc[-1, 1]
others_res = sheep.wool(df, raw_data, days=days)
# 去除st
history = tool.history_name(start_date=df['trade_date'].min())
history['name'] = 'st'
df = df.merge(history, on=['ts_code', 'trade_date'], how='left')
# print(df.shape)
df = df[df['name'].isna()]
df.drop(columns=['name'], inplace=True)
# print(daily_basic.shape)
# print(df.shape)
# df.to_csv(path+'data.csv')
# 1.10天上涨大于等于7次的股票
# 2.所有股票筛选100*(ma1/ma10-1)在[1,10]之间
# 3.筛选换手率,市值
df = df[df['count_%s' % LABEL] >= TIMES]
# print(df.shape)
df = df[(df['ma1/ma10pct'] >= up_n_pct[0])
& (df['ma1/ma10pct'] <= up_n_pct[1])]
# print(df.shape)
df = df.merge(daily_basic[['ts_code', 'trade_date']],
on=['ts_code', 'trade_date'])
# print(df.shape)
df.to_csv(path + '%daysup%stimes.csv' % (PERIOD, TIMES))
sheep.wool(df, data).to_csv(path + '%daysup%stimeshuisu.csv' % (PERIOD, TIMES))
# df[df['trade_date']==(str(today).replace('-',''))][['ts_code']].reset_index(drop=True).to_csv('%sdaysup%stimes%snew.txt'%(PERIOD,TIMES,str(today)),index=False)
print('10day up 6',
df[df['trade_date'] == (str(today).replace('-', ''))].shape[0])
on=['ts_code', 'trade_date'],
how='left')
stock_data1 = stock_data1[stock_data1['name'].isna()]
stock_data1.drop(columns='name', inplace=True)
df = pd.DataFrame()
for day in stock_data1['trade_date'].unique():
# df=pd.concat([data[data['trade_date']==day].sort_values(by='trade_date',ascending=False).head(30),df])
df = pd.concat([
stock_data1[stock_data1['trade_date'] == day].sort_values(
by='score', ascending=False).head(30), df
])
# stock.to_csv(path + str((mv_bins[i], mv_bins[i + 1])) + "30ofall_marks.csv")
df.to_csv(path + str((mv_bins[i], mv_bins[i + 1])) +
"-30-of-bins.csv")
stock = sheep.wool(df, data, days=days)
if stock.iloc[-1, -1] > 1:
print(stock.iloc[-1, -1])
stock.to_csv(path + str((mv_bins[i], mv_bins[i + 1])) +
"pct-wool.csv")
# print(stock)
#所有股票排名回溯
top_n = [50]
for it in top_n:
for switch in [False]:
df = pd.DataFrame()
stock_mark = stock_mark.merge(history_name,
on=['ts_code', 'trade_date'],
how='left')
stock_mark = stock_mark[stock_mark['name'].isna()]
def first_up(
start_date='20200201',
end_date='20200215',
PRICEB='open',
PRICES='close',
sell_date=1,
buy_date=2,
CUTS=2,
CHANGE=['open', 'pre_close'],
):
NOTCONTAIN = sfilter.StockFilter().stock_basic(end_date,
name="st|ST",
market="科创板")
raw_data = pd.read_sql_query(
'select * from daily where (trade_date>=%(start)s and trade_date<=%(end)s)',
params={
'start': start_date,
'end': end_date
},
con=engine)
stk_limit = pd.read_sql_query(
'select * from stk_limit where (trade_date>=%(start)s and trade_date<=%(end)s)',
params={
'start': start_date,
'end': end_date
},
con=engine)
raw_data.drop_duplicates(inplace=True)
stk_limit.drop_duplicates(inplace=True)
print(raw_data.shape, stk_limit.shape)
raw_data = raw_data[raw_data["ts_code"].isin(NOTCONTAIN['ts_code']) ==
False]
print('科创,st,out', raw_data.shape)
raw_data = raw_data.merge(
stk_limit.loc[:, ['ts_code', 'trade_date', 'up_limit', 'down_limit']],
on=['ts_code', 'trade_date'])
raw_data['%s/%s' % (CHANGE[0], CHANGE[1])] = raw_data.apply(
lambda x: 99 if x[CHANGE[0]] == x['up_limit'] else -99
if x[CHANGE[0]] == x['down_limit'] else x[CHANGE[0]] / x[CHANGE[1]],
axis=1)
if buy_date != 1:
price_buy = basic.basic().pre_data(
raw_data[[
'ts_code', 'trade_date', PRICEB,
'%s/%s' % (CHANGE[0], CHANGE[1])
]],
label=[PRICEB, '%s/%s' % (CHANGE[0], CHANGE[1])],
new_label=[
'price_buy',
'xishu_pct',
],
pre_days=-buy_date + 1)
raw_data = raw_data.merge(
price_buy[['ts_code', 'trade_date', 'price_buy', 'xishu_pct']],
on=['ts_code', 'trade_date'])
print(raw_data.shape)
raw_data.dropna(inplace=True)
print(raw_data.shape)
# 筛选收盘涨停数据
data = raw_data.loc[:, [
'ts_code', 'trade_date', 'close', 'pre_close', 'up_limit', 'xishu_pct'
]]
print(data.shape)
data = data.merge(basic.basic().pre_data(
data[['ts_code', 'trade_date', 'up_limit']],
label=['up_limit'],
new_label=['pre_up_limit'])[['ts_code', 'trade_date', 'pre_up_limit']],
on=['ts_code', 'trade_date'])
print(data.shape)
data = data[data['close'] == data['up_limit']]
# print(data.shape)
# raw_data=raw_data[raw_data["ts_code"].isin(data['ts_code'].unique()) == True]
# print('不在涨停df中的股票out',raw_data.shape)
# print(raw_data['ts_code'].unique().shape)
# buy_data = raw_data.loc[:, ['ts_code', 'trade_date', PRICES]]
# z1=sheep.wool(data, raw_data, PRICEB='price_buy')
# print(z1.iloc[-1, -1])
z2 = sheep.wool(data, raw_data)
print(z2.iloc[-1, -1])
# t1=sheep.wool(data[data['pre_close'] != data['pre_up_limit']],
# raw_data, PRICEB='price_buy')
# print(t1.iloc[-1, -1])
t2 = sheep.wool(data[data['pre_close'] != data['pre_up_limit']], raw_data)
print(t2.iloc[-1, -1])
res_all = pd.DataFrame()
res_some = pd.DataFrame()
# data.to_csv('%s-%slimit_stock_list.csv' % (start_date, end_date))
for cut in range(-10, 12, CUTS):
if cut < -9:
buy_data = data[data['xishu_pct'] == -99]
print('open down')
elif cut > 10:
buy_data = data[data['xishu_pct'] == 99]
print('open up')
else:
buy_data = data[(data['xishu_pct'] < (1 + 0.01 * cut))
& (data['xishu_pct'] >= (0.98 + 0.01 * cut))]
print('%s<=x<%s' % ((0.98 + 0.01 * cut), (1 + 0.01 * cut)))
if buy_data.empty:
res_all.loc[cut, 'res'] = None
res_some.loc[cut, 'res'] = None
print(cut, buy_data['xishu_pct'].max(),
buy_data['xishu_pct'].min(), buy_data['xishu_pct'].mean())
continue
wool_all = sheep.wool(buy_data,
raw_data,
PRICEB='price_buy',
days=sell_date)
if wool_all.iloc[-1, -1]:
# wool_all.to_csv('%s-%spct%swool_all.csv' % (start_date, end_date, cut))
print('all%s' % cut, wool_all.iloc[-1, -1])
res_all.loc[cut, 'res'] = wool_all.iloc[-1, -1]
res_all.loc[cut, 'n_mean'] = wool_all['n'].mean()
res_all.loc[cut, 'days'] = wool_all.shape[0]
wool_some = sheep.wool(
buy_data[buy_data['pre_close'] != buy_data['pre_up_limit']],
raw_data,
PRICEB='price_buy',
days=sell_date)
if wool_some.iloc[-1, -1]:
# wool_some.to_csv('%s-%spct%swool_some.csv' % (start_date, end_date, cut))
print('some', wool_some.iloc[-1, -1])
res_some.loc[cut, 'res'] = wool_some.iloc[-1, -1]
res_some.loc[cut, 'n_mean'] = wool_some['n'].mean()
res_some.loc[cut, 'days'] = wool_some.shape[0]