How to use close method in chromeless

1# coding:utf-82#3# The MIT License (MIT)4#5# Copyright (c) 2016-2021 yutiansut/QUANTAXIS6#7# Permission is hereby granted, free of charge, to any person obtaining a copy8# of this software and associated documentation files (the "Software"), to deal9# in the Software without restriction, including without limitation the rights10# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell11# copies of the Software, and to permit persons to whom the Software is12# furnished to do so, subject to the following conditions:13#14# The above copyright notice and this permission notice shall be included in all15# copies or substantial portions of the Software.16#17# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR18# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,19# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE20# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER21# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,22# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE23# SOFTWARE.24from functools import reduce25import numpy as np26import pandas as pd27from QUANTAXIS.QAIndicator.base import *28"""29DataFrame 类30以下的函数都可以被直接add_func31"""32"""331.	趋向指标 34又叫趋势跟踪类指标,主要用于跟踪并预测股价的发展趋势35包含的主要指标361. 移动平均线 MA372. 指数平滑移动平均线 MACD383. 趋向指标 DMI394. 瀑布线 PBX405. 平均线差 DMA416. 动力指标(动量线)  MTM427. 指数平均线 EXPMA438. 佳庆指标 CHO44"""45def QA_indicator_MA(DataFrame,*args,**kwargs):46    """MA47    48    Arguments:49        DataFrame {[type]} -- [description]50    51    Returns:52        [type] -- [description]53    """54    CLOSE = DataFrame['close']55    return pd.DataFrame({'MA{}'.format(N): MA(CLOSE, N)  for N in list(args)})56def QA_indicator_MA_VOL(DataFrame,*args,**kwargs):57    """MA_VOLU58    59    Arguments:60        DataFrame {[type]} -- [description]61    62    Returns:63        [type] -- [description]64    """65    VOL = DataFrame['volume']66    return pd.DataFrame({'MA_VOL{}'.format(N): MA(VOL, N)  for N in list(args)})67def QA_indicator_EMA(DataFrame, N):68    CLOSE = DataFrame['close']69    return pd.DataFrame({'EMA': EMA(CLOSE, N)})70def QA_indicator_SMA(DataFrame, N):71    CLOSE = DataFrame['close']72    return pd.DataFrame({'SMA': SMA(CLOSE, N)})73def QA_indicator_MACD(DataFrame, short=12, long=26, mid=9):74    """75    MACD CALC76    """77    CLOSE = DataFrame['close']78    DIF = EMA(CLOSE, short)-EMA(CLOSE, long)79    DEA = EMA(DIF, mid)80    MACD = (DIF-DEA)*281    return pd.DataFrame({'DIF': DIF, 'DEA': DEA, 'MACD': MACD})82def QA_indicator_DMI(DataFrame, M1=14, M2=6):83    """84    趋向指标 DMI85    """86    HIGH = DataFrame.high87    LOW = DataFrame.low88    CLOSE = DataFrame.close89    OPEN = DataFrame.open90    TR = SUM(MAX(MAX(HIGH-LOW, ABS(HIGH-REF(CLOSE, 1))),91                 ABS(LOW-REF(CLOSE, 1))), M1)92    HD = HIGH-REF(HIGH, 1)93    LD = REF(LOW, 1)-LOW94    DMP = SUM(IFAND(HD>0,HD>LD,HD,0), M1)95    DMM = SUM(IFAND(LD>0,LD>HD,LD,0), M1)96    DI1 = DMP*100/TR97    DI2 = DMM*100/TR98    ADX = MA(ABS(DI2-DI1)/(DI1+DI2)*100, M2)99    ADXR = (ADX+REF(ADX, M2))/2100    return pd.DataFrame({101        'DI1': DI1, 'DI2': DI2,102        'ADX': ADX, 'ADXR': ADXR103    })104def QA_indicator_PBX(DataFrame, N1=3, N2=5, N3=8, N4=13, N5=18, N6=24):105    '瀑布线'106    C = DataFrame['close']107    PBX1 = (EMA(C, N1) + EMA(C, 2 * N1) + EMA(C, 4 * N1)) / 3108    PBX2 = (EMA(C, N2) + EMA(C, 2 * N2) + EMA(C, 4 * N2)) / 3109    PBX3 = (EMA(C, N3) + EMA(C, 2 * N3) + EMA(C, 4 * N3)) / 3110    PBX4 = (EMA(C, N4) + EMA(C, 2 * N4) + EMA(C, 4 * N4)) / 3111    PBX5 = (EMA(C, N5) + EMA(C, 2 * N5) + EMA(C, 4 * N5)) / 3112    PBX6 = (EMA(C, N6) + EMA(C, 2 * N6) + EMA(C, 4 * N6)) / 3113    DICT = {'PBX1': PBX1, 'PBX2': PBX2, 'PBX3': PBX3,114            'PBX4': PBX4, 'PBX5': PBX5, 'PBX6': PBX6}115    return pd.DataFrame(DICT)116def QA_indicator_DMA(DataFrame, M1=10, M2=50, M3=10):117    """118    平均线差 DMA119    """120    CLOSE = DataFrame.close121    DDD = MA(CLOSE, M1) - MA(CLOSE, M2)122    AMA = MA(DDD, M3)123    return pd.DataFrame({124        'DDD': DDD, 'AMA': AMA125    })126def QA_indicator_MTM(DataFrame, N=12, M=6):127    '动量线'128    C = DataFrame.close129    mtm = C - REF(C, N)130    MTMMA = MA(mtm, M)131    DICT = {'MTM': mtm, 'MTMMA': MTMMA}132    return pd.DataFrame(DICT)133def QA_indicator_EXPMA(DataFrame, P1=5, P2=10, P3=20, P4=60):134    """ 指数平均线 EXPMA"""135    CLOSE = DataFrame.close136    MA1 = EMA(CLOSE, P1)137    MA2 = EMA(CLOSE, P2)138    MA3 = EMA(CLOSE, P3)139    MA4 = EMA(CLOSE, P4)140    return pd.DataFrame({141        'MA1': MA1, 'MA2': MA2, 'MA3': MA3, 'MA4': MA4142    })143def QA_indicator_CHO(DataFrame, N1=10, N2=20, M=6):144    """145    佳庆指标 CHO146    """147    HIGH = DataFrame.high148    LOW = DataFrame.low149    CLOSE = DataFrame.close150    VOL = DataFrame.volume151    MID = SUM(VOL*(2*CLOSE-HIGH-LOW)/(HIGH+LOW), 0)152    CHO = MA(MID, N1)-MA(MID, N2)153    MACHO = MA(CHO, M)154    return pd.DataFrame({155        'CHO': CHO, 'MACHO': MACHO156    })157"""1582.	反趋向指标159主要捕捉趋势的转折点160随机指标KDJ161乖离率 BIAS162变动速率 ROC163顺势指标 CCI164威廉指标 W&R165震荡量(变动速率) OSC166相对强弱指标 RSI167动态买卖指标 ADTM168"""169def QA_indicator_KDJ(DataFrame, N=9, M1=3, M2=3):170    C = DataFrame['close']171    H = DataFrame['high']172    L = DataFrame['low']173    RSV = ((C - LLV(L, N)) / (HHV(H, N) - LLV(L, N)) * 100).groupby('code').fillna(method='ffill')174    K = SMA(RSV, M1)175    D = SMA(K, M2)176    J = 3 * K - 2 * D177    DICT = {'KDJ_K': K, 'KDJ_D': D, 'KDJ_J': J}178    return pd.DataFrame(DICT)179def QA_indicator_BIAS(DataFrame, N1, N2, N3):180    '乖离率'181    CLOSE = DataFrame['close']182    BIAS1 = (CLOSE - MA(CLOSE, N1)) / MA(CLOSE, N1) * 100183    BIAS2 = (CLOSE - MA(CLOSE, N2)) / MA(CLOSE, N2) * 100184    BIAS3 = (CLOSE - MA(CLOSE, N3)) / MA(CLOSE, N3) * 100185    DICT = {'BIAS1': BIAS1, 'BIAS2': BIAS2, 'BIAS3': BIAS3}186    return pd.DataFrame(DICT)187def QA_indicator_ROC(DataFrame, N=12, M=6):188    '变动率指标'189    C = DataFrame['close']190    roc = 100 * (C - REF(C, N)) / REF(C, N)191    ROCMA = MA(roc, M)192    DICT = {'ROC': roc, 'ROCMA': ROCMA}193    return pd.DataFrame(DICT)194def QA_indicator_CCI(DataFrame, N=14):195    """196    TYP:=(HIGH+LOW+CLOSE)/3;197    CCI:(TYP-MA(TYP,N))/(0.015*AVEDEV(TYP,N));198    """199    typ = (DataFrame['high'] + DataFrame['low'] + DataFrame['close']) / 3200    ## 此处AVEDEV可能为0值  因此导致出错 +0.0000000000001201    cci = ((typ - MA(typ, N)) / (0.015 * AVEDEV(typ, N) + 0.00000001))202    a = 100203    b = -100204    return pd.DataFrame({205        'CCI': cci, 'a': a, 'b': b206    })207def QA_indicator_WR(DataFrame, N, N1):208    '威廉指标'209    HIGH = DataFrame['high']210    LOW = DataFrame['low']211    CLOSE = DataFrame['close']212    WR1 = 100 * (HHV(HIGH, N) - CLOSE) / (HHV(HIGH, N) - LLV(LOW, N))213    WR2 = 100 * (HHV(HIGH, N1) - CLOSE) / (HHV(HIGH, N1) - LLV(LOW, N1))214    DICT = {'WR1': WR1, 'WR2': WR2}215    return pd.DataFrame(DICT)216def QA_indicator_OSC(DataFrame, N=20, M=6):217    """变动速率线218    震荡量指标OSC，也叫变动速率线。属于超买超卖类指标,是从移动平均线原理派生出来的一种分析指标。219    它反应当日收盘价与一段时间内平均收盘价的差离值,从而测出股价的震荡幅度。220    按照移动平均线原理，根据OSC的值可推断价格的趋势，如果远离平均线，就很可能向平均线回归。221    """222    C = DataFrame['close']223    OS = (C - MA(C, N)) * 100224    MAOSC = EMA(OS, M)225    DICT = {'OSC': OS, 'MAOSC': MAOSC}226    return pd.DataFrame(DICT)227def QA_indicator_RSI(DataFrame, N1=12, N2=26, N3=9):228    '相对强弱指标RSI1:SMA(MAX(CLOSE-LC,0),N1,1)/SMA(ABS(CLOSE-LC),N1,1)*100;'229    CLOSE = DataFrame['close']230    LC = REF(CLOSE, 1)231    RSI1 = SMA(MAX(CLOSE - LC, 0), N1) / SMA(ABS(CLOSE - LC), N1) * 100232    RSI2 = SMA(MAX(CLOSE - LC, 0), N2) / SMA(ABS(CLOSE - LC), N2) * 100233    RSI3 = SMA(MAX(CLOSE - LC, 0), N3) / SMA(ABS(CLOSE - LC), N3) * 100234    DICT = {'RSI1': RSI1, 'RSI2': RSI2, 'RSI3': RSI3}235    return pd.DataFrame(DICT)236def QA_indicator_ADTM(DataFrame, N=23, M=8):237    '动态买卖气指标'238    HIGH = DataFrame.high239    LOW = DataFrame.low240    OPEN = DataFrame.open241    DTM = IF(OPEN > REF(OPEN, 1), MAX((HIGH - OPEN), (OPEN - REF(OPEN, 1))), 0)242    DBM = IF(OPEN < REF(OPEN, 1), MAX((OPEN - LOW), (OPEN - REF(OPEN, 1))), 0)243    STM = SUM(DTM, N)244    SBM = SUM(DBM, N)245    ADTM1 = IF(STM > SBM, (STM - SBM) / STM,246               IF(STM != SBM, (STM - SBM) / SBM, 0))247    MAADTM = MA(ADTM1, M)248    DICT = {'ADTM': ADTM1, 'MAADTM': MAADTM}249    return pd.DataFrame(DICT)250"""2513.	量能指标252通过成交量的大小和变化研判趋势变化253容量指标 VR254量相对强弱 VRSI255能量指标 CR256人气意愿指标 ARBR257成交量标准差 VSTD"""258def QA_indicator_VR(DataFrame, M1=26, M2=100, M3=200):259    VOL = DataFrame.volume260    CLOSE = DataFrame.close261    LC = REF(CLOSE, 1)262    VR = SUM(IF(CLOSE > LC, VOL, 0), M1)/SUM(IF(CLOSE <= LC, VOL, 0), M1)*100263    a = M2264    b = M3265    return pd.DataFrame({266        'VR': VR, 'a': a, 'b': b267    })268def QA_indicator_VRSI(DataFrame, N=6):269    VOL = DataFrame.volume270    vrsi = SMA(MAX(VOL-REF(VOL, 1), 0), N, 1) / \271        SMA(ABS(VOL-REF(VOL, 1)), N, 1)*100272    return pd.DataFrame({'VRSI': vrsi})273def QA_indicator_CR(DataFrame, N=26, M1=5, M2=10, M3=20):274    HIGH = DataFrame.high275    LOW = DataFrame.low276    CLOSE = DataFrame.close277    VOL = DataFrame.volume278    MID = (HIGH+LOW+CLOSE)/3279    CR = SUM(MAX(0, HIGH-REF(MID, 1)), N)/SUM(MAX(0, REF(MID, 1)-LOW), N)*100280    MA1 = REF(MA(CR, M1), M1/2.5+1)281    MA2 = REF(MA(CR, M2), M2/2.5+1)282    MA3 = REF(MA(CR, M3), M3/2.5+1)283    return pd.DataFrame({284        'CR': CR, 'MA1': MA1, 'MA2': MA2, 'MA3': MA3285    })286def QA_indicator_ARBR(DataFrame, M1=26, M2=70, M3=150):287    HIGH = DataFrame.high288    LOW = DataFrame.low289    CLOSE = DataFrame.close290    OPEN = DataFrame.open291    AR = SUM(HIGH-OPEN, M1)/SUM(OPEN-LOW, M1)*100292    BR = SUM(MAX(0, HIGH-REF(CLOSE, 1)), M1) / \293        SUM(MAX(0, REF(CLOSE, 1)-LOW), M1)*100294    a = M2295    b = M3296    return pd.DataFrame({297        'AR': AR, 'BR': BR, 'a': a, 'b': b298    })299def QA_indicator_VSTD(DataFrame, N=10):300    VOL = DataFrame.volume301    vstd = STD(VOL, N)302    return pd.DataFrame({'VSTD': vstd})303"""3044.	量价指标305通过成交量和股价变动关系分析未来趋势306震荡升降指标ASI307价量趋势PVT308能量潮OBV309量价趋势VPT310"""311def QA_indicator_ASI(DataFrame, M1=26, M2=10):312    """313    LC=REF(CLOSE,1);314    AA=ABS(HIGH-LC);315    BB=ABS(LOW-LC);316    CC=ABS(HIGH-REF(LOW,1));317    DD=ABS(LC-REF(OPEN,1));318    R=IF(AA>BB AND AA>CC,AA+BB/2+DD/4,IF(BB>CC AND BB>AA,BB+AA/2+DD/4,CC+DD/4));319    X=(CLOSE-LC+(CLOSE-OPEN)/2+LC-REF(OPEN,1));320    SI=16*X/R*MAX(AA,BB);321    ASI:SUM(SI,M1);322    ASIT:MA(ASI,M2);323    """324    CLOSE = DataFrame['close']325    HIGH = DataFrame['high']326    LOW = DataFrame['low']327    OPEN = DataFrame['open']328    LC = REF(CLOSE, 1)329    AA = ABS(HIGH - LC)330    BB = ABS(LOW-LC)331    CC = ABS(HIGH - REF(LOW, 1))332    DD = ABS(LC - REF(OPEN, 1))333    R = IFAND(AA > BB, AA > CC, AA+BB/2+DD/4,334           IFAND(BB > CC, BB > AA, BB+AA/2+DD/4, CC+DD/4))335    X = (CLOSE - LC + (CLOSE - OPEN) / 2 + LC - REF(OPEN, 1))336    SI = 16*X/R*MAX(AA, BB)337    ASI = SUM(SI, M1)338    ASIT = MA(ASI, M2)339    return pd.DataFrame({340        'ASI': ASI, 'ASIT': ASIT341    })342def QA_indicator_PVT(DataFrame):343    CLOSE = DataFrame.close344    VOL = DataFrame.volume345    PVT = SUM((CLOSE-REF(CLOSE, 1))/REF(CLOSE, 1)*VOL, 0)346    return pd.DataFrame({'PVT': PVT})347def QA_indicator_OBV(DataFrame):348    """能量潮"""349    VOL = DataFrame.volume350    CLOSE = DataFrame.close351    return pd.DataFrame({352        'OBV': np.cumsum(IF(CLOSE > REF(CLOSE, 1), VOL, IF(CLOSE < REF(CLOSE, 1), -VOL, 0)))/10000353    })354def QA_indicator_VPT(DataFrame, N=51, M=6):355    VOL = DataFrame.volume356    CLOSE = DataFrame.close357    VPT = SUM(VOL*(CLOSE-REF(CLOSE, 1))/REF(CLOSE, 1), 0)358    MAVPT = MA(VPT, M)359    return pd.DataFrame({360        'VPT': VPT, 'MAVPT': MAVPT361    })362"""3635.	压力支撑指标364主要用于分析股价目前收到的压力和支撑365布林带 BOLL366麦克指标 MIKE367"""368def QA_indicator_BOLL(DataFrame, N=20, P=2):369    '布林线'370    C = DataFrame['close']371    boll = MA(C, N)372    UB = boll + P * STD(C, N)373    LB = boll - P * STD(C, N)374    DICT = {'BOLL': boll, 'UB': UB, 'LB': LB}375    return pd.DataFrame(DICT)376def QA_indicator_MIKE(DataFrame, N=12):377    """378    MIKE指标379    指标说明380    MIKE是另外一种形式的路径指标。381    买卖原则382    1  WEAK-S，MEDIUM-S，STRONG-S三条线代表初级、中级、强力支撑。383    2  WEAK-R，MEDIUM-R，STRONG-R三条线代表初级、中级、强力压力。384    """385    HIGH = DataFrame.high386    LOW = DataFrame.low387    CLOSE = DataFrame.close388    TYP = (HIGH+LOW+CLOSE)/3389    LL = LLV(LOW, N)390    HH = HHV(HIGH, N)391    WR = TYP+(TYP-LL)392    MR = TYP+(HH-LL)393    SR = 2*HH-LL394    WS = TYP-(HH-TYP)395    MS = TYP-(HH-LL)396    SS = 2*LL-HH397    return pd.DataFrame({398        'WR': WR, 'MR': MR, 'SR': SR,399        'WS': WS, 'MS': MS, 'SS': SS400    })401def QA_indicator_BBI(DataFrame, N1=3, N2=6, N3=12, N4=24):402    '多空指标'403    C = DataFrame['close']404    bbi = (MA(C, N1) + MA(C, N2) + MA(C, N3) + MA(C, N4)) / 4405    DICT = {'BBI': bbi}406    return pd.DataFrame(DICT)407def QA_indicator_MFI(DataFrame, N=14):408    """409    资金指标410    TYP := (HIGH + LOW + CLOSE)/3;411    V1:=SUM(IF(TYP>REF(TYP,1),TYP*VOL,0),N)/SUM(IF(TYP<REF(TYP,1),TYP*VOL,0),N);412    MFI:100-(100/(1+V1));413    赋值: (最高价 + 最低价 + 收盘价)/3414    V1赋值:如果TYP>1日前的TYP,返回TYP*成交量(手),否则返回0的N日累和/如果TYP<1日前的TYP,返回TYP*成交量(手),否则返回0的N日累和415    输出资金流量指标:100-(100/(1+V1))416    """417    C = DataFrame['close']418    H = DataFrame['high']419    L = DataFrame['low']420    VOL = DataFrame['volume']421    TYP = (C + H + L) / 3422    V1 = SUM(IF(TYP > REF(TYP, 1), TYP * VOL, 0), N) / \423        SUM(IF(TYP < REF(TYP, 1), TYP * VOL, 0), N)424    mfi = 100 - (100 / (1 + V1))425    DICT = {'MFI': mfi}426    return pd.DataFrame(DICT)427def QA_indicator_ATR(DataFrame, N=14):428    """429    输出TR:(最高价-最低价)和昨收-最高价的绝对值的较大值和昨收-最低价的绝对值的较大值430    输出真实波幅:TR的N日简单移动平均431    算法：今日振幅、今日最高与昨收差价、今日最低与昨收差价中的最大值，为真实波幅，求真实波幅的N日移动平均432    参数：N　天数，一般取14433    """434    C = DataFrame['close']435    H = DataFrame['high']436    L = DataFrame['low']437    TR = MAX(MAX((H - L), ABS(REF(C, 1) - H)), ABS(REF(C, 1) - L))438    atr = MA(TR, N)439    return pd.DataFrame({'TR': TR, 'ATR': atr})440def QA_indicator_SKDJ(DataFrame, N=9, M=3):441    """442    1.指标>80 时，回档机率大；指标<20 时，反弹机率大；443    2.K在20左右向上交叉D时，视为买进信号参考； 444    3.K在80左右向下交叉D时，视为卖出信号参考；445    4.SKDJ波动于50左右的任何讯号，其作用不大。446    """447    CLOSE = DataFrame['close']448    LOWV = LLV(DataFrame['low'], N)449    HIGHV = HHV(DataFrame['high'], N)450    RSV = EMA((CLOSE - LOWV) / (HIGHV - LOWV) * 100, M)451    K = EMA(RSV, M)452    D = MA(K, M)453    DICT = {'RSV': RSV, 'SKDJ_K': K, 'SKDJ_D': D}454    return pd.DataFrame(DICT)455def QA_indicator_DDI(DataFrame, N=13, N1=26, M=1, M1=5):456    """457    '方向标准离差指数'458    分析DDI柱状线，由红变绿(正变负)，卖出信号参考；由绿变红，买入信号参考。459    """460    H = DataFrame['high']461    L = DataFrame['low']462    DMZ = IF((H + L) > (REF(H, 1) + REF(L, 1)), 463             MAX(ABS(H - REF(H, 1)), ABS(L - REF(L, 1))), 0)464    DMF = IF((H + L) < (REF(H, 1) + REF(L, 1)),465             MAX(ABS(H - REF(H, 1)), ABS(L - REF(L, 1))), 0)466    DIZ = SUM(DMZ, N) / (SUM(DMZ, N) + SUM(DMF, N))467    DIF = SUM(DMF, N) / (SUM(DMF, N) + SUM(DMZ, N))468    ddi = DIZ - DIF469    ADDI = SMA(ddi, N1, M)470    AD = MA(ADDI, M1)471    DICT = {'DDI': ddi, 'ADDI': ADDI, 'AD': AD}472    return pd.DataFrame(DICT)473def QA_indicator_shadow(DataFrame):474    """475    上下影线指标476    """477    return {478        'LOW': lower_shadow(DataFrame), 'UP': upper_shadow(DataFrame),479        'BODY': body(DataFrame), 'BODY_ABS': body_abs(DataFrame), 'PRICE_PCG': price_pcg(DataFrame)480    }481def lower_shadow(DataFrame):  # 下影线482    return abs(DataFrame.low - MIN(DataFrame.open, DataFrame.close))483def upper_shadow(DataFrame):  # 上影线484    return abs(DataFrame.high - MAX(DataFrame.open, DataFrame.close))485def body_abs(DataFrame):486    return abs(DataFrame.open - DataFrame.close)487def body(DataFrame):488    return DataFrame.close - DataFrame.open489def price_pcg(DataFrame):490    return body(DataFrame) / DataFrame.open491def amplitude(DataFrame):492    return (DataFrame.high - DataFrame.low) / DataFrame.low493"""4946.	大盘指标495通过涨跌家数研究大盘指数的走势496涨跌比率 ADR497绝对幅度指标 ABI498新三价率 TBR499腾落指数 ADL500广量冲力指标501指数平滑广量 STIX...

...34    def set_mon_open(self, open):35        self.__mon_open = open36    def get_mon_open(self):37        return self.__mon_open38    def set_mon_close(self, close):39        self.__mon_close = close40    def get_mon_close(self):41        return self.__mon_close42    def set_mon_break_start(self, bstart):43        self.__mon_bstart = bstart44    def get_mon_break_start(self):45        return self.__mon_bstart46    def set_mon_break_end(self, bend):47        self.__mon_bend = bend48    def get_mon_break_end(self):49        return self.__mon_bend50    def set_tues_open(self, open):51        self.__tues_open = open52    def get_tues_open(self):53        return self.__tues_open54    def set_tues_close(self, close):55        self.__tues_close = close56    def get_tues_close(self):57        return self.__tues_close58    def set_tues_break_start(self, bstart):59        self.__tues_bstart = bstart60    def get_tues_break_start(self):61        return self.__tues_bstart62    def set_tues_break_end(self, bend):63        self.__tues_bend = bend64    def get_tues_break_end(self):65        return self.__tues_bend66    def set_wed_open(self, open):67        self.__wed_open = open68    def get_wed_open(self):69        return self.__wed_open70    def set_wed_close(self, close):71        self.__wed_close = close72    def get_wed_close(self):73        return self.__wed_close74    def set_wed_break_start(self, bstart):75        self.__wed_bstart = bstart76    def get_wed_break_start(self):77        return self.__wed_bstart78    def set_wed_break_end(self, bend):79        self.__wed_bend = bend80    def get_wed_break_end(self):81        return self.__wed_bend82    def set_thur_open(self, open):83        self.__thur_open = open84    def get_thur_open(self):85        return self.__thur_open86    def set_thur_close(self, close):87        self.__thur_close = close88    def get_thur_close(self):89        return self.__thur_close90    def set_thur_break_start(self, bstart):91        self.__thur_bstart = bstart92    def get_thur_break_start(self):93        return self.__thur_bstart94    def set_thur_break_end(self, bend):95        self.__thur_bend = bend96    def get_thur_break_end(self):97        return self.__thur_bend98    def set_fri_open(self, open):99        self.__fri_open = open100    def get_fri_open(self):101        return self.__fri_open102    def set_fri_close(self, close):103        self.__fri_close = close104    def get_fri_close(self):105        return self.__fri_close106    def set_fri_break_start(self, bstart):107        self.__fri_bstart = bstart108    def get_fri_break_start(self):109        return self.__fri_bstart110    def set_fri_break_end(self, bend):111        self.__fri_bend = bend112    def get_fri_break_end(self):113        return self.__fri_bend114    def set_sat_open(self, open):115        self.__sat_open = open116    def get_sat_open(self):117        return self.__sat_open118    def set_sat_close(self, close):119        self.__sat_close = close120    def get_sat_close(self):121        return self.__sat_close122    def set_sat_break_start(self, bstart):123        self.__sat_bstart = bstart124    def get_sat_break_start(self):125        return self.__sat_bstart126    def set_sat_break_end(self, bend):127        self.__sat_bend = bend128    def get_sat_break_end(self):129        return self.__sat_bend130    def set_sun_open(self, open):131        self.__sun_open = open132    def get_sun_open(self):133        return self.__sun_open134    def set_sun_close(self, close):135        self.__sun_close = close136    def get_sun_close(self):137        return self.__sun_close138    def set_sun_break_start(self, bstart):139        self.__sun_bstart = bstart140    def get_sun_break_start(self):141        return self.__sun_bstart142    def set_sun_break_end(self, bend):143        self.__sun_bend = bend144    def get_sun_break_end(self):145        return self.__sun_bend146OperatingHour.__init__(self, mon_open, mon_close, mon_bstart, mon_bend, tues_open, tues_close, tues_bstart,147                               tues_bend, wed_open, wed_close, wed_bstart, wed_bend, thur_open, thur_close, thur_bstart,148                               thur_bend, fri_open, fri_close, fri_bstart, fri_bend, sat_open, sat_close, sat_bstart,149                               sat_bend, sun_open, sun_close, sun_bstart, sun_bend)150mon_open = TimeField('Opening Time:', [validators.Optional()])151mon_close = TimeField('Closing Time:', [validators.Optional()])152mon_break_start = TimeField('Break Start Time:', [validators.Optional()])153mon_break_end = TimeField('Break End Time:', [validators.Optional()])154tues_open = TimeField('Opening Time:', [validators.Optional()])155tues_close = TimeField('Closing Time:', [validators.Optional()])156tues_break_start = TimeField('Break Start Time:', [validators.Optional()])157tues_break_end = TimeField('Break End Time:', [validators.Optional()])158wed_open = TimeField('Opening Time:', [validators.Optional()])159wed_close = TimeField('Closing Time:', [validators.Optional()])160wed_break_start = TimeField('Break Start Time:', [validators.Optional()])161wed_break_end = TimeField('Break End Time:', [validators.Optional()])162thur_open = TimeField('Opening Time:', [validators.Optional()])163thur_close = TimeField('Closing Time:', [validators.Optional()])164thur_break_start = TimeField('Break Start Time:', [validators.Optional()])165thur_break_end = TimeField('Break End Time:', [validators.Optional()])166fri_open = TimeField('Opening Time:', [validators.Optional()])167fri_close = TimeField('Closing Time:', [validators.Optional()])168fri_break_start = TimeField('Break Start Time:', [validators.Optional()])169fri_break_end = TimeField('Break End Time:', [validators.Optional()])170sat_open = TimeField('Opening Time:', [validators.Optional()])171sat_close = TimeField('Closing Time:', [validators.Optional()])172sat_break_start = TimeField('Break Start Time:', [validators.Optional()])173sat_break_end = TimeField('Break End Time:', [validators.Optional()])174sun_open = TimeField('Opening Time:', [validators.Optional()])175sun_close = TimeField('Closing Time:', [validators.Optional()])176sun_break_start = TimeField('Break Start Time:', [validators.Optional()])177sun_break_end = TimeField('Break End Time:', [validators.Optional()])178def edit_clinic_info():179    # create_user_form = ClinicInfoForm(request.form)180    # if request.method == 'POST' and create_user_form.validate():181    #     users_dict = {}182    #     db = shelve.open('clinic_storage.db', 'c')183    #184    #     try:185    #         users_dict = db['Users']186    #     except:187    #         print("Error in retrieving Users from storage.db.")188    #189    #     user = User.ClinicInfo(create_user_form.mon_open.data, create_user_form.mon_close.data, create_user_form.mon_break_start.data, create_user_form.mon_break_end.data,190    #                            create_user_form.tues_open.data, create_user_form.tues_close.data, create_user_form.tues_break_start.data, create_user_form.tues_break_end.data,191    #                            create_user_form.wed_open.data, create_user_form.wed_close.data, create_user_form.wed_break_start.data, create_user_form.wed_break_end.data,192    #                            create_user_form.thur_open.data, create_user_form.thur_close.data, create_user_form.thur_break_start.data, create_user_form.thur_break_end.data,193    #                            create_user_form.fri_open.data, create_user_form.fri_close.data, create_user_form.fri_break_start.data, create_user_form.fri_break_end.data,194    #                            create_user_form.sat_open.data, create_user_form.sat_close.data, create_user_form.sat_break_start.data, create_user_form.sat_break_end.data,195    #                            create_user_form.sun_open.data, create_user_form.sun_close.data, create_user_form.sun_break_start.data, create_user_form.sun_break_end.data,196    #                            create_user_form.off_start.data, create_user_form.off_end.data, create_user_form.off_reason.data, create_user_form.mon_break_end.data,197    #                            create_user_form.password.data, create_user_form.cfm_password.data, create_user_form.name.data,198    #                            create_user_form.address.data, create_user_form.email.data, create_user_form.phone.data)199    #     users_dict[user.get_clinic_id()] = user200    #     db['Users'] = users_dict201    #202    #     # Test codes203    #     users_dict = db['Users']204    #     user = users_dict[user.get_clinic_id()]205    #     print(user.get_name(), "was stored in storage.db successfully with Clinic ID =", user.get_clinic_id())206    #207    #     db.close()208    #     return redirect(url_for('login_clinic'))...

1# -*- coding:utf-8 -*-2# editor: gmj3# Date: 2019-12-03 14:144# desc: 辅助方法集合5def cut_short_period(index_list):6    # 除去序号值差值小于4的区间7    while 1:8        max_index = len(index_list) - 19        # 只有一个值  或者以0开头的区间分割index列表10        if max_index == 0 or (index_list[0] == 0 and max_index == 1):11            break12        for k, v in enumerate(index_list):13            if k < max_index:14                if k == 0:15                    if index_list[1] - index_list[0] < 3:16                        del index_list[k + 1]17                    continue18                if index_list[k + 1] - index_list[k] < 4:19                    if k == max_index - 1:20                        # 如果是第一段数据少于4，则只删除index_list[k + 1]21                        del index_list[k + 1]22                    else:23                        del index_list[k + 1]24                        del index_list[k]25                    break26        if max_index + 1 == len(index_list):27            break28def deviation_judge(section_1: tuple, section_2: tuple = None):29    """30    根据价格 和 MACD  kdj_D  来判断是否背离31    以及判断振幅和量能的异常32    :param section_1: 离现在最近的区间  (DateFrame数据,index 列表)33    :param section_2: 倒数第三个区间34    :return:  背离状态35    """36    deviation_status = {}37    if section_2:38        data_df1, peak_index1 = section_139        up_sign = 1 if data_df1['MACD'].iloc[peak_index1[0]] > 0 else -140        section_1_close_macd = (data_df1['close'].iloc[peak_index1[0]], data_df1['MACD'].iloc[peak_index1[0]],41                                data_df1['kdj_D'].iloc[peak_index1[0]])42        data_df3, peak_index3 = section_243        section_2_close_macd = (data_df3['close'].iloc[peak_index3[0]], data_df3['MACD'].iloc[peak_index3[0]],44                                data_df3['kdj_D'].iloc[peak_index3[0]])45        if len(peak_index3) > 1:46            # 判断出最大的macd值的index47            close_macd_list = []48            macd_list = []49            for i in peak_index3:50                close_macd_list.append((data_df3['close'].iloc[i], data_df3['MACD'].iloc[i], data_df3['kdj_D'].iloc[i]))51                macd_list.append(abs(data_df3['MACD'].iloc[i]))52            max_macd = max(macd_list)53            for close_macd in close_macd_list:54                if close_macd[1] == max_macd:55                    section_2_close_macd = close_macd56        # 判断背离57        if (section_1_close_macd[0] - section_2_close_macd[0]) * (58                section_1_close_macd[1] - section_2_close_macd[1]) < 0:59            if up_sign > 0:60                deviation_status['macd_deviation_status'] = '上涨MACD背离'61            else:62                deviation_status['macd_deviation_status'] = '下跌MACD背离'63        if (section_1_close_macd[0] - section_2_close_macd[0]) * (64                section_1_close_macd[2] - section_2_close_macd[2]) < 0:65            if up_sign > 0:66                deviation_status['kdj_deviation_status'] = '上涨KDJ背离'67            else:68                deviation_status['kdj_deviation_status'] = '下跌KDJ背离'69    else:70        data_df1, peak_index = section_171        up_sign = 1 if data_df1['MACD'].iloc[peak_index[0]] > 0 else -172        if (data_df1['close'].iloc[peak_index[0]] - data_df1['close'].iloc[peak_index[1]]) * \73                (data_df1['MACD'].iloc[peak_index[0]] - data_df1['MACD'].iloc[peak_index[1]]) < 0:74            if up_sign > 0:75                deviation_status['macd_deviation_status'] = '上涨MACD背离'76            else:77                deviation_status['macd_deviation_status'] = '下跌MACD背离'78        if (data_df1['close'].iloc[peak_index[0]] - data_df1['close'].iloc[peak_index[1]]) * \79                (data_df1['kdj_D'].iloc[peak_index[0]] - data_df1['kdj_D'].iloc[peak_index[1]]) < 0:80            if up_sign > 0:81                deviation_status['kdj_deviation_status'] = '上涨KDJ背离'82            else:83                deviation_status['kdj_deviation_status'] = '下跌KDJ背离'84    # 量能异常 和 振幅异常提示85    warning_res = swing_vol_warning(data_df1)86    if warning_res:87        deviation_status['swing_vol_warning'] = warning_res88    # deviation_status['macd_trend']='MACD 上涨趋势' if up_sign else 'MACD 下跌趋势'89    return deviation_status90def swing_vol_warning(data_df):91    # 量能异常 和 振幅异常提示92    for row in range(data_df.shape[0]):93        row_data = data_df.iloc[row]94        if (row_data['vol'] / row_data['vol_MA_10']) >= 2 and row_data['swing'] > 0.05:95            return f"{row_data['trade_date']}-振幅和量能异常"96        else:97            return None98def get_peak_index(macd_section):99    """100    :param macd_section: 一个macd指标值区间的DataFrame101    :return:  该区间的极值index102    """103    # macd_section=macd_sec['MACD']104    peak_index = []105    # 判断符号106    # macd_section['MACD'].sum()107    for s_i in range(1, macd_section['MACD'].shape[0] - 1):108        # 判断极值 如果 该值与两侧值的差值乘积符号为正，则为极值109        if (macd_section['MACD'].iloc[s_i] - macd_section['MACD'].iloc[s_i - 1]) * (110                macd_section['MACD'].iloc[s_i] - macd_section['MACD'].iloc[s_i + 1]) > 0:111            # 上升找极大值  下跌找极小值112            if (macd_section['MACD'].iloc[s_i] - macd_section['MACD'].iloc[s_i - 1]) * macd_section[113                'MACD'].sum() > 0:  # 剔除凹值114                peak_index.append(s_i)115    # 第一区间的值还在持续变化，那么极值就是第一值116    if not peak_index:117        peak_index.append(0)118    cut_short_period(peak_index)119    return peak_index120def deviation_macd_judge(section_1: tuple, section_2: tuple = None):121    """122    根据价格 和 MACD   来判断是否背离123    以及判断振幅和量能的异常124    :param section_1: 离现在最近的区间  (DateFrame数据,index 列表)125    :param section_2: 倒数第三个区间126    :return:  背离状态127    """128    deviation_status = {}129    if section_2:130        data_df1, peak_index1 = section_1131        up_sign = 1 if data_df1['MACD'].iloc[peak_index1[0]] > 0 else -1132        section_1_close_macd = (data_df1['close'].iloc[peak_index1[0]], data_df1['MACD'].iloc[peak_index1[0]])133        data_df3, peak_index3 = section_2134        section_2_close_macd = (data_df3['close'].iloc[peak_index3[0]], data_df3['MACD'].iloc[peak_index3[0]])135        if len(peak_index3) > 1:136            # 判断出最大的macd值的index137            close_macd_list = []138            macd_list = []139            for i in peak_index3:140                close_macd_list.append((data_df3['close'].iloc[i], data_df3['MACD'].iloc[i], data_df3['kdj_D'].iloc[i]))141                macd_list.append(abs(data_df3['MACD'].iloc[i]))142            max_macd = max(macd_list)143            for close_macd in close_macd_list:144                if close_macd[1] == max_macd:145                    section_2_close_macd = close_macd146        # 判断背离147        if (section_1_close_macd[0] - section_2_close_macd[0]) * (148                section_1_close_macd[1] - section_2_close_macd[1]) < 0:149            if up_sign > 0:150                deviation_status['macd_deviation_status'] = '上涨MACD背离'151            else:152                deviation_status['macd_deviation_status'] = '下跌MACD背离'153    else:154        data_df1, peak_index = section_1155        up_sign = 1 if data_df1['MACD'].iloc[peak_index[0]] > 0 else -1156        if (data_df1['close'].iloc[peak_index[0]] - data_df1['close'].iloc[peak_index[1]]) * \157                (data_df1['MACD'].iloc[peak_index[0]] - data_df1['MACD'].iloc[peak_index[1]]) < 0:158            if up_sign > 0:159                deviation_status['macd_deviation_status'] = '上涨MACD背离'160            else:161                deviation_status['macd_deviation_status'] = '下跌MACD背离'162    # 量能异常 和 振幅异常提示163    warning_res = swing_vol_warning(data_df1)164    if warning_res:165        deviation_status['swing_vol_warning'] = warning_res...

...32        self.assertRaises(db.DBNoSuchFileError, d2.open,33                self.filename+"2", db.DB_BTREE, db.DB_THREAD, 0666)34        d.put("test","this is a test")35        self.assertEqual(d.get("test"), "this is a test", "put!=get")36        dbenv.close()  # This "close" should close the child db handle also37        self.assertRaises(db.DBError, d.get, "test")38    def test02_close_dbenv_before_dbcursor(self):39        dbenv = db.DBEnv()40        dbenv.open(self.homeDir,41                   db.DB_INIT_CDB| db.DB_CREATE |db.DB_THREAD|db.DB_INIT_MPOOL,42                   0666)43        d = db.DB(dbenv)44        d.open(self.filename, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)45        d.put("test","this is a test")46        d.put("test2","another test")47        d.put("test3","another one")48        self.assertEqual(d.get("test"), "this is a test", "put!=get")49        c=d.cursor()50        c.first()51        c.next()52        d.close()  # This "close" should close the child db handle also53     # db.close should close the child cursor54        self.assertRaises(db.DBError,c.next)55        d = db.DB(dbenv)56        d.open(self.filename, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)57        c=d.cursor()58        c.first()59        c.next()60        dbenv.close()61    # The "close" should close the child db handle also, with cursors62        self.assertRaises(db.DBError, c.next)63    def test03_close_db_before_dbcursor_without_env(self):64        import os.path65        path=os.path.join(self.homeDir,self.filename)66        d = db.DB()67        d.open(path, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)68        d.put("test","this is a test")69        d.put("test2","another test")70        d.put("test3","another one")71        self.assertEqual(d.get("test"), "this is a test", "put!=get")72        c=d.cursor()73        c.first()74        c.next()75        d.close()76    # The "close" should close the child db handle also77        self.assertRaises(db.DBError, c.next)78    def test04_close_massive(self):79        dbenv = db.DBEnv()80        dbenv.open(self.homeDir,81                   db.DB_INIT_CDB| db.DB_CREATE |db.DB_THREAD|db.DB_INIT_MPOOL,82                   0666)83        dbs=[db.DB(dbenv) for i in xrange(16)]84        cursors=[]85        for i in dbs :86            i.open(self.filename, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)87        dbs[10].put("test","this is a test")88        dbs[10].put("test2","another test")89        dbs[10].put("test3","another one")90        self.assertEqual(dbs[4].get("test"), "this is a test", "put!=get")91        for i in dbs :92            cursors.extend([i.cursor() for j in xrange(32)])93        for i in dbs[::3] :94            i.close()95        for i in cursors[::3] :96            i.close()97    # Check for missing exception in DB! (after DB close)98        self.assertRaises(db.DBError, dbs[9].get, "test")99    # Check for missing exception in DBCursor! (after DB close)100        self.assertRaises(db.DBError, cursors[101].first)101        cursors[80].first()102        cursors[80].next()103        dbenv.close()  # This "close" should close the child db handle also104    # Check for missing exception! (after DBEnv close)105        self.assertRaises(db.DBError, cursors[80].next)106    def test05_close_dbenv_delete_db_success(self):107        dbenv = db.DBEnv()108        dbenv.open(self.homeDir,109                   db.DB_INIT_CDB| db.DB_CREATE |db.DB_THREAD|db.DB_INIT_MPOOL,110                   0666)111        d = db.DB(dbenv)112        d.open(self.filename, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)113        dbenv.close()  # This "close" should close the child db handle also114        del d115        try:116            import gc117        except ImportError:118            gc = None119        if gc:120            # force d.__del__ [DB_dealloc] to be called121            gc.collect()122    def test06_close_txn_before_dup_cursor(self) :123        dbenv = db.DBEnv()124        dbenv.open(self.homeDir,db.DB_INIT_TXN | db.DB_INIT_MPOOL |125                db.DB_INIT_LOG | db.DB_CREATE)126        d = db.DB(dbenv)127        txn = dbenv.txn_begin()128        d.open(self.filename, dbtype = db.DB_HASH, flags = db.DB_CREATE,129                txn=txn)130        d.put("XXX", "yyy", txn=txn)131        txn.commit()132        txn = dbenv.txn_begin()133        c1 = d.cursor(txn)134        c2 = c1.dup()135        self.assertEqual(("XXX", "yyy"), c1.first())136        # Not interested in warnings about implicit close.137        import warnings138        if sys.version_info < (2, 6) :139            # Completely resetting the warning state is140            # problematic with python >=2.6 with -3 (py3k warning),141            # because some stdlib modules selectively ignores warnings.142            warnings.simplefilter("ignore")143            txn.commit()144            warnings.resetwarnings()145        else :146            # When we drop support for python 2.4147            # we could use: (in 2.5 we need a __future__ statement)148            #149            #    with warnings.catch_warnings():150            #        warnings.simplefilter("ignore")151            #        txn.commit()152            #153            # We can not use "with" as is, because it would be invalid syntax154            # in python 2.4 and (with no __future__) 2.5.155            # Here we simulate "with" following PEP 343 :156            w = warnings.catch_warnings()157            w.__enter__()158            try :159                warnings.simplefilter("ignore")160                txn.commit()161            finally :162                w.__exit__()163        self.assertRaises(db.DBCursorClosedError, c2.first)164    def test07_close_db_before_sequence(self):165        import os.path166        path=os.path.join(self.homeDir,self.filename)167        d = db.DB()168        d.open(path, db.DB_BTREE, db.DB_CREATE | db.DB_THREAD, 0666)169        dbs=db.DBSequence(d)170        d.close()  # This "close" should close the child DBSequence also171        dbs.close()  # If not closed, core dump (in Berkeley DB 4.6.*)172#----------------------------------------------------------------------173def test_suite():174    suite = unittest.TestSuite()175    suite.addTest(unittest.makeSuite(DBEnvClosedEarlyCrash))176    return suite177if __name__ == '__main__':...

1import pandas as pd2import numpy as np3COLUMNS_CHART_DATA = ['date', 'open', 'high', 'low', 'close', 'volume']4COLUMNS_TRAINING_DATA_V1 = [5    'open_lastclose_ratio', 'high_close_ratio', 'low_close_ratio',6    'close_lastclose_ratio', 'volume_lastvolume_ratio',7    'close_ma5_ratio', 'volume_ma5_ratio',8    'close_ma10_ratio', 'volume_ma10_ratio',9    'close_ma20_ratio', 'volume_ma20_ratio',10    'close_ma60_ratio', 'volume_ma60_ratio',11    'close_ma120_ratio', 'volume_ma120_ratio',12]13COLUMNS_TRAINING_DATA_V1_RICH = [14    'open_lastclose_ratio', 'high_close_ratio', 'low_close_ratio',15    'close_lastclose_ratio', 'volume_lastvolume_ratio',16    'close_ma5_ratio', 'volume_ma5_ratio',17    'close_ma10_ratio', 'volume_ma10_ratio',18    'close_ma20_ratio', 'volume_ma20_ratio',19    'close_ma60_ratio', 'volume_ma60_ratio',20    'close_ma120_ratio', 'volume_ma120_ratio',21    'inst_lastinst_ratio', 'frgn_lastfrgn_ratio',22    'inst_ma5_ratio', 'frgn_ma5_ratio',23    'inst_ma10_ratio', 'frgn_ma10_ratio',24    'inst_ma20_ratio', 'frgn_ma20_ratio',25    'inst_ma60_ratio', 'frgn_ma60_ratio',26    'inst_ma120_ratio', 'frgn_ma120_ratio',27]28COLUMNS_TRAINING_DATA_V2 = [29    'per', 'pbr', 'roe',30    'open_lastclose_ratio', 'high_close_ratio', 'low_close_ratio',31    'close_lastclose_ratio', 'volume_lastvolume_ratio',32    'close_ma5_ratio', 'volume_ma5_ratio',33    'close_ma10_ratio', 'volume_ma10_ratio',34    'close_ma20_ratio', 'volume_ma20_ratio',35    'close_ma60_ratio', 'volume_ma60_ratio',36    'close_ma120_ratio', 'volume_ma120_ratio',37    'market_kospi_ma5_ratio', 'market_kospi_ma20_ratio', 38    'market_kospi_ma60_ratio', 'market_kospi_ma120_ratio', 39    'bond_k3y_ma5_ratio', 'bond_k3y_ma20_ratio', 40    'bond_k3y_ma60_ratio', 'bond_k3y_ma120_ratio'41]42def preprocess(data, ver='v1'):43    windows = [5, 10, 20, 60, 120]44    for window in windows:45        data['close_ma{}'.format(window)] = \46            data['close'].rolling(window).mean()47        data['volume_ma{}'.format(window)] = \48            data['volume'].rolling(window).mean()49        data['close_ma%d_ratio' % window] = \50            (data['close'] - data['close_ma%d' % window]) \51            / data['close_ma%d' % window]52        data['volume_ma%d_ratio' % window] = \53            (data['volume'] - data['volume_ma%d' % window]) \54            / data['volume_ma%d' % window]55            56        if ver == 'v1.rich':57            data['inst_ma{}'.format(window)] = \58                data['close'].rolling(window).mean()59            data['frgn_ma{}'.format(window)] = \60                data['volume'].rolling(window).mean()61            data['inst_ma%d_ratio' % window] = \62                (data['close'] - data['inst_ma%d' % window]) \63                / data['inst_ma%d' % window]64            data['frgn_ma%d_ratio' % window] = \65                (data['volume'] - data['frgn_ma%d' % window]) \66                / data['frgn_ma%d' % window]67    data['open_lastclose_ratio'] = np.zeros(len(data))68    data.loc[1:, 'open_lastclose_ratio'] = \69        (data['open'][1:].values - data['close'][:-1].values) \70        / data['close'][:-1].values71    data['high_close_ratio'] = \72        (data['high'].values - data['close'].values) \73        / data['close'].values74    data['low_close_ratio'] = \75        (data['low'].values - data['close'].values) \76        / data['close'].values77    data['close_lastclose_ratio'] = np.zeros(len(data))78    data.loc[1:, 'close_lastclose_ratio'] = \79        (data['close'][1:].values - data['close'][:-1].values) \80        / data['close'][:-1].values81    data['volume_lastvolume_ratio'] = np.zeros(len(data))82    data.loc[1:, 'volume_lastvolume_ratio'] = \83        (data['volume'][1:].values - data['volume'][:-1].values) \84        / data['volume'][:-1] \85            .replace(to_replace=0, method='ffill') \86            .replace(to_replace=0, method='bfill').values87    if ver == 'v1.rich':88        data['inst_lastinst_ratio'] = np.zeros(len(data))89        data.loc[1:, 'inst_lastinst_ratio'] = \90            (data['inst'][1:].values - data['inst'][:-1].values) \91            / data['inst'][:-1] \92                .replace(to_replace=0, method='ffill') \93                .replace(to_replace=0, method='bfill').values94        data['frgn_lastfrgn_ratio'] = np.zeros(len(data))95        data.loc[1:, 'frgn_lastfrgn_ratio'] = \96            (data['frgn'][1:].values - data['frgn'][:-1].values) \97            / data['frgn'][:-1] \98                .replace(to_replace=0, method='ffill') \99                .replace(to_replace=0, method='bfill').values100    return data101def load_data(fpath, date_from, date_to, ver='v1'):102    header = None if ver == 'v1' else 0103    data = pd.read_csv(fpath, thousands=',', header=header, 104        converters={'date': lambda x: str(x)})105    if ver == 'v1':106        data.columns = ['date', 'open', 'high', 'low', 'close', 'volume']107    # 날짜 오름차순 정렬108    data = data.sort_values(by='date').reset_index()109    # 데이터 전처리110    data = preprocess(data)111    112    # 기간 필터링113    data['date'] = data['date'].str.replace('-', '')114    data = data[(data['date'] >= date_from) & (data['date'] <= date_to)]115    data = data.dropna()116    # 차트 데이터 분리117    chart_data = data[COLUMNS_CHART_DATA]118    # 학습 데이터 분리119    training_data = None120    if ver == 'v1':121        training_data = data[COLUMNS_TRAINING_DATA_V1]122    elif ver == 'v1.rich':123        training_data = data[COLUMNS_TRAINING_DATA_V1_RICH]124    elif ver == 'v2':125        data.loc[:, ['per', 'pbr', 'roe']] = \126            data[['per', 'pbr', 'roe']].apply(lambda x: x / 100)127        training_data = data[COLUMNS_TRAINING_DATA_V2]128        training_data = training_data.apply(np.tanh)129    else:130        raise Exception('Invalid version.')131    ...

1import numpy as np2import pandas as pd 3#设定基础函数4def get_delta(x,N):5    lst=[]6    for i in range(1,N+2):7        lst.append(x)8        x=x.diff()9    return lst[-1]10def rank(array):11    s = pd.Series(array)12    return s.rank(ascending=False)[len(s)-1]13#单只股票的time series分析指标14def get_alpha1(df):15    df['a1']=(df['high']-df['low'])/abs(df['close']-df['open'])16    return df['a1']17def get_alpha2(df):18    df['a2']=(df['high']-df['close'])/(df['close']-df['low'])19    return df['a2']20def get_alpha3(df):21    df['a3']=(df['high']-df['open'])/(df['open']-df['close'])22    return df['a3']23# Alpha#1: (SignedPower(((returns < 0) ? stddev(returns, 20) : close), 2.)24def get_alpha4(df):25    filter4_1=df['returns'].rolling(window=20).std().where(df['returns']<0)26    df['ralpha4']=filter4_1.fillna(df['close'])27    return df['ralpha4']**228#Alpha#3:correlation(high,volume,5)29def get_alpha5(df):30    return df['high'].rolling(window=5).corr(df['volume'])31#Alpha#4:correlation(open,volume,10)32def get_alpha6(df):33    return df['open'].rolling(window=10).corr(df['volume'])34    35#Alpha#5:delta(sum(open,5)*sum(return,5),10)36def get_alpha7(df):37    return get_delta(df['open'].rolling(window=5).sum()*df['returns'].rolling(window=5).sum(),10)38#Alpha#9: ((0 < ts_min(delta(close, 1), 5)) ? delta(close, 1) : ((ts_max(delta(close, 1), 5) < 0) ? delta(close, 1) : (-1 * delta(close, 1)))) 39def get_alpha8(df):40    filter9_1=df['close'].diff().where((df['close'].diff().rolling(window=5).min())<0)41    filter9_2=filter9_1.fillna(df['close'].diff().where((df['close'].diff().rolling(window=5).max())>0))42    return filter9_2.fillna(-df['close'].diff())43#Alpha#12: (delta(volume, 1)) * (-1 * delta(close, 1))44def get_alpha9(df):45    return -1*df['volume'].diff()*df['close'].diff()46#Alpha#18: stddev(abs((close - open)), 5) + (close - open)) + correlation(close, open, 10)47def get_alpha10(df):48    return abs(df['close']-df['open']).rolling(window=5).std()+(df['close']-df['open'])+get_alpha7(df)49#Alpha#23: (((sum(high, 20) / 20) < high) ? (-1 * delta(high, 2)) : 0) 50def get_alpha11(df):51    df['alpha12']=-1*get_delta(df['high'],2).where(df['high'].rolling(window=20).mean()<df['high'])52    return df['alpha12'].fillna(0)53#Alpha#24: ((((delta((sum(close, 100) / 100), 100) / delay(close, 100)) < 0.05) || ((delta((sum(close, 100) / 100), 100) / delay(close, 100)) == 0.05)) ? (-1 * (close - ts_min(close, 100))) : (-1 * delta(close, 3))) 54def get_alpha12(df):55    nofilter13=df['close'].rolling(window=100).min()-df['close']56    filter13_1=nofilter13.where(get_delta(df['close'].rolling(window=100).mean(),100)/(df['close']-get_delta(df['close'],100))<0.05)57    filter13_2=filter13_1.fillna(nofilter13.where(get_delta(df['close'].rolling(window=100).mean(),100)/(df['close']-get_delta(df['close'],100))==0.05))58    return filter13_2.fillna(-get_delta(df['close'],3))59#Alpha#26: (-1 * ts_max(correlation(ts_rank(volume, 5), ts_rank(high, 5), 5), 3)) 60def get_alpha13(df):61    return df['volume'].rolling(window=5).apply(func=rank).rolling(window=5).corr(df['high'].rolling(window=5).apply(func=rank)).rolling(window=3).max()62#Alpha#35: ((Ts_Rank(volume, 32) * (1 - Ts_Rank(((close + high) - low), 16))) * (1 - Ts_Rank(returns, 32))) 63def get_alpha14(df):64    return df['volume'].rolling(window=32).apply(func=rank)*(1-(df['close']+df['high']-df['low']).rolling(window=32).apply(func=rank))*(1-df['returns'].rolling(window=32).apply(func=rank))65#Alpha#43: (ts_rank((volume / adv20), 20) * ts_rank((-1 * delta(close, 7)), 8))  66def get_alpha15(df):67    return df['volume']/df['close']*df['volume'].rolling(window=20).mean().rolling(window=20).apply(func=rank)*-get_delta(df['close'],7).rolling(window=8).apply(func=rank)68#Alpha#55: correlation((close - ts_min(low, 12)) / (ts_max(high, 12) - ts_min(low, 12))), volume, 6)69def get_alpha16(df):70    return df['volume'].rolling(window=6).corr((df['close']-df['low'].rolling(window=12).min())/(df['high'].rolling(window=12).max()-df['low'].rolling(window=12).min()))71#Alpha#101: ((close - open) / ((high - low) + .001)) 72def get_alpha17(df):73    return (df['close']-df['open'])/(df['high']-df['low']+0.01)74#Alpha#28: (correlation(adv20, low, 5) + ((high + low) / 2)) - close75def get_alpha18(df):76    return df['low'].rolling(window=5).corr(df['close']*df['volume'].rolling(window=20).mean())+(df['high']+df['low'])/2-df['close']77#EMV:A=（今日最高+今日最低）/2;B=（前日最高+前日最低）/2;C=今日最高-今日最低;2.EM=（A-B）*C/今日成交额;3.EMV=N日内EM的累和;4.MAEMV=EMV的M日简单移动平均.参数N为14，参数M为978def get_alpha19(df):79    return get_delta((df['high']+df['low'])/2,1).rolling(window=14).sum().rolling(window=9).mean()80#MTMMA动量指标81def get_alpha20(df):82    return get_delta(df['close'],12).rolling(window=6).mean()83#横截面分析指标84#Alpha#41: (((high * low)^0.5) - vwap) 85#Alpha#25: rank(((((-1 * returns) * adv20) * vwap) * (high - close)))  86#Alpha#84: SignedPower(Ts_Rank((vwap - ts_max(vwap, 15.3217)), 20.7127), delta(close, 4.96796)) ...

...23            },24            'write' : {25                'call' : 'evaluate',26                'write' : """open("%(path)s", 'ab+').write(__import__("base64").urlsafe_b64decode('%(chunk_b64)s'))""",27                'truncate' : """open("%(path)s", 'w').close()"""28            },29            'read' : {30                'call': 'evaluate',31                'read' : """__import__("base64").b64encode(open("%(path)s", "rb").read())"""32            },33            'md5' : {34                'call': 'evaluate',35                'md5': """__import__("hashlib").md5(open("%(path)s", 'rb').read()).hexdigest()"""36            },37            'evaluate' : {38                'call': 'render',39                'evaluate': """%(code)s""",40                'test_os': """'-'.join([__import__('os').name, __import__('sys').platform])""",41                'test_os_expected': '^[\w-]+$'...

1import pandas as pd2import numpy as np3import keras4import tensorflow as tf5from keras.preprocessing.sequence import TimeseriesGenerator6import plotly.graph_objects as go7#Implementation8filename = "GOOG.csv"9df = pd.read_csv(filename)10print(df.info())11df['Date'] = pd.to_datetime(df['Date'])12df.set_axis(df['Date'], inplace=True)13df.drop(columns=['Open', 'High', 'Low', 'Volume'], inplace=True)14#Data Preprocessing15close_data = df['Close'].values16close_data = close_data.reshape((-1,1))17split_percent = 0.8018split = int(split_percent * len(close_data))19close_train = close_data[:split]20close_test = close_data[split:]21date_train = df['Date'][:split]22date_test = df['Date'][split:]23print(len(close_train))24print(len(close_test))25look_back = 1526train_generator = TimeseriesGenerator(close_train, close_train, length=look_back, batch_size=20)     27test_generator = TimeseriesGenerator(close_test, close_test, length=look_back, batch_size=1)28#Neural Network29from keras.models import Sequential30from keras.layers import LSTM, Dense31model = Sequential()32model.add(33    LSTM(10,34        activation = 'relu',35        input_shape=(look_back,1))36)37model.add(Dense(1))38model.compile(optimizer = 'adam', loss = 'mse')39num_epochs = 2540model.fit_generator(train_generator, epochs = num_epochs, verbose = 1)41prediction = model.predict_generator(test_generator)42close_train = close_train.reshape((-1))43close_test = close_test.reshape((-1))44prediction = prediction.reshape((-1))45#print("Predicted:", [float(f"{x:.1f}") for x in prediction])46trace1 = go.Scatter(47    x = date_train,48    y = close_train,49    mode = 'lines',50    name = 'Data'51)52trace2 = go.Scatter(53    x = date_test,54    y = prediction,55    mode = 'lines',56    name = 'Prediction'57)58trace3 = go.Scatter(59    x = date_test,60    y = close_test,61    mode='lines',62    name = 'Ground Truth'63)64layout = go.Layout(65    title = "Google Stock",66    xaxis = {'title' : "Date"},67    yaxis = {'title' : "Close"}68)69fig = go.Figure(data=[trace1, trace2, trace3], layout=layout)70fig.show()71close_data = close_data.reshape((-1))72def predict(num_prediction, model):73    prediction_list = close_data[-look_back:]74    75    for _ in range(num_prediction):76        x = prediction_list[-look_back:]77        x = x.reshape((1, look_back, 1))78        out = model.predict(x)[0][0]79        prediction_list = np.append(prediction_list, out)80    prediction_list = prediction_list[look_back-1:]81        82    return prediction_list83    84def predict_dates(num_prediction):85    last_date = df['Date'].values[-1]86    prediction_dates = pd.date_range(last_date, periods=num_prediction+1).tolist()87    return prediction_dates88num_prediction = 3089forecast = predict(num_prediction, model)...