def _genSignalIDs(self, idt, original_ids, signal_type):
IDs = original_ids
FilterLevel = 0
for i in range(self.FiltrationLevel):
iArgs = self["第"+str(i)+"层"]
if iArgs.SignalType!=signal_type: continue
if iArgs.IDFilter:
iIDs = self._FT.getFilteredID(idt, id_filter_str=iArgs.IDFilter)
IDs = sorted(set(iIDs).intersection(set(IDs)))
if iArgs.GroupFactors:
GroupData = self._FT.readData(dts=[idt], ids=IDs, factor_names=list(iArgs.GroupFactors)).iloc[:,0,:]
if GroupData.shape[0]>0: GroupData[pd.isnull(GroupData)] = np.nan
AllGroups = [GroupData[iGroup].unique().tolist() for iGroup in iArgs.GroupFactors]
AllGroups = CartesianProduct(AllGroups)
IDs = []
for jGroup in AllGroups:
jMask = pd.Series(True, index=GroupData.index)
for k, kSubGroup in enumerate(jGroup):
if pd.notnull(kSubGroup): jMask = (jMask & (GroupData[iArgs.GroupFactors[k]]==kSubGroup))
else: jMask = (jMask & pd.isnull(GroupData[iArgs.GroupFactors[k]]))
jIDs = self._filtrateID(idt, GroupData[jMask].index.tolist(), iArgs)
IDs += jIDs
else:
IDs = self._filtrateID(idt, IDs, iArgs)
FilterLevel += 1
if FilterLevel>0: return IDs
else: return []
def __QS_prepareRawData__(self, factor_names, ids, dts, args={}):
FactorInfo = self._FactorDB._FactorInfo.loc[self.Name]
Fields = [self._IDField]+factor_names
DBFields = FactorInfo["DBFieldName"].loc[Fields].tolist()
DBTableName = self._FactorDB._TableInfo.loc[self.Name, "DBTableName"]
RawData = pd.DataFrame(columns=DBFields)
FieldStr = ",".join(DBFields)
ConditionField = FactorInfo[pd.notnull(FactorInfo["Supplementary"])]
if ConditionField.shape[0]>0:
SingleCondition, MultiCondition = {}, OrderedDict()
for i, iCondition in enumerate(ConditionField.index):
iConditionValue = args.get(iCondition, self[iCondition])
if iConditionValue=="All": MultiCondition[ConditionField["DBFieldName"].iloc[i]] = ConditionField["Supplementary"].iloc[i].split(",")[1:]
else: SingleCondition[ConditionField["DBFieldName"].iloc[i]] = iConditionValue
if MultiCondition:
RawData = None
MultiCondition, MultiConditionValue = list(MultiCondition.keys()), CartesianProduct(list(MultiCondition.values()))
for iMultiConditionValue in MultiConditionValue:
SingleCondition.update(dict(zip(MultiCondition, iMultiConditionValue)))
if RawData is None: RawData = self._FactorDB._ts.query(DBTableName, fields=FieldStr, **SingleCondition)
else: RawData = RawData.append(self._FactorDB._ts.query(DBTableName, fields=FieldStr, **SingleCondition))
else: RawData = self._FactorDB._ts.query(DBTableName, fields=FieldStr, **SingleCondition)
else:
RawData = self._FactorDB._ts.query(DBTableName, fields=FieldStr)
RawData = RawData.loc[:, DBFields]
RawData.columns = ["ID"]+factor_names
RawData["ID"] = self._FactorDB.DBID2ID(RawData["ID"])
return RawData
def _allocateWeight(self, idt, ids, original_ids, args):
nID = len(ids)
if not args.GroupFactors:# 没有类别因子
if args.WeightFactor=='等权': NewSignal = pd.Series(1/nID, index=ids)
else:
WeightData = self._FT.readData(factor_names=[args.WeightFactor], dts=[idt], ids=ids).iloc[0,0,:]
if args.WeightMiss=='舍弃': WeightData = WeightData[pd.notnull(WeightData)]
else: WeightData[pd.notnull(WeightData)] = WeightData.mean()
WeightData = WeightData / WeightData.sum()
NewSignal = WeightData
else:
GroupData = self._FT.readData(factor_names=args.GroupFactors, dts=[idt], ids=original_ids).iloc[:,0,:]
GroupData[pd.isnull(GroupData)] = np.nan
AllGroups = [GroupData[iGroup].unique().tolist() for iGroup in args.GroupFactors]
AllGroups = CartesianProduct(AllGroups)
nGroup = len(AllGroups)
if args.GroupWeight=='等权': GroupWeight = pd.Series(np.ones(nGroup)/nGroup, dtype='float')
else:
GroupWeight = pd.Series(index=np.arange(nGroup), dtype='float')
GroupWeightData = self._FT.readData(factor_names=[args.GroupWeight], dts=[idt], ids=original_ids).iloc[0,0,:]
for i, iGroup in enumerate(AllGroups):
if pd.notnull(iGroup[0]): iMask = (GroupData[args.GroupFactors[0]]==iGroup[0])
else: iMask = pd.isnull(GroupData[args.GroupFactors[0]])
for j, jSubGroup in enumerate(iGroup[1:]):
if pd.notnull(jSubGroup): iMask = (iMask & (GroupData[args.GroupFactors[j+1]]==jSubGroup))
else: iMask = (iMask & pd.isnull(GroupData[args.GroupFactors[j+1]]))
GroupWeight.iloc[i] = GroupWeightData[iMask].sum()
GroupWeight[pd.isnull(GroupWeight)] = 0
GroupTotalWeight = GroupWeight.sum()
if GroupTotalWeight!=0: GroupWeight = GroupWeight/GroupTotalWeight
if args.WeightFactor=='等权': WeightData = pd.Series(1.0, index=original_ids)
else: WeightData = self._FT.readData(factor_names=[args.WeightFactor], dts=[idt], ids=original_ids).iloc[0,0,:]
SelectedGroupData = GroupData.loc[ids]
NewSignal = pd.Series()
for i, iGroup in enumerate(AllGroups):
if pd.notnull(iGroup[0]): iMask = (SelectedGroupData[args.GroupFactors[0]]==iGroup[0])
else: iMask = pd.isnull(SelectedGroupData[args.GroupFactors[0]])
for j, jSubGroup in enumerate(iGroup[1:]):
if pd.notnull(jSubGroup): iMask = (iMask & (SelectedGroupData[args.GroupFactors[j+1]]==jSubGroup))
else: iMask = (iMask & pd.isnull(SelectedGroupData[args.GroupFactors[j+1]]))
iIDs = SelectedGroupData[iMask].index.tolist()
if (iIDs==[]) and (args.GroupMiss=='全配'):
if pd.notnull(iGroup[0]): iMask = (GroupData[args.GroupFactors[0]]==iGroup[0])
else: iMask = pd.isnull(GroupData[args.GroupFactors[0]])
for k, kSubClass in enumerate(iGroup[1:]):
if pd.notnull(kSubClass): iMask = (iMask & (GroupData[args.GroupFactors[k+1]]==kSubClass))
else: iMask = (iMask & pd.isnull(GroupData[args.GroupFactors[k+1]]))
iIDs = GroupData[iMask].index.tolist()
elif (iIDs==[]) and (args.GroupMiss=='忽略'): continue
iSignal = WeightData.loc[iIDs]
iSignalWeight = iSignal.sum()
if iSignalWeight!=0: iSignal = iSignal / iSignalWeight * GroupWeight.iloc[i]
else: iSignal = iSignal*0.0
if args.WeightMiss=='填充均值': iSignal[pd.isnull(iSignal)] = iSignal.mean()
NewSignal = NewSignal.append(iSignal[pd.notnull(iSignal) & (iSignal!=0)])
NewSignal = NewSignal / NewSignal.sum()
return NewSignal
def changeMultiClass2SingleClass(multi_class, sep=None):
MultiClass = []
for i in range(multi_class.shape[1]):
MultiClass.append(pd.unique(multi_class[:, i]).tolist())
MultiClass = CartesianProduct(MultiClass)
SingleClassData = np.empty(shape=(multi_class.shape[0], ), dtype="O")
ClassDict = {}
for i, iMultiClass in enumerate(MultiClass):
iMask = np.array([True] * multi_class.shape[0])
if sep is not None:
iSingleClass = sep.join(map(str, iMultiClass))
else:
iSingleClass = str(i)
for j, jSubClass in enumerate(iMultiClass):
if pd.notnull(jSubClass):
iMask = iMask & (multi_class[:, j] == jSubClass)
else:
iMask = iMask & pd.isnull(multi_class[:, j])
SingleClassData[iMask] = iSingleClass
ClassDict[iSingleClass] = iMultiClass
return (SingleClassData, ClassDict)