def test21():
print("Simulation: ST ContextDHondtINF")
jobID: str = "Roulette1"
selector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
#pProbToolOLin0802HLin1002:APenalization = PenalizationToolDefinition.exportProbPenaltyToolOStat08HLin1002(
# InputSimulatorDefinition.numberOfAggrItems)
pToolOLin0802HLin1002: APenalization = PenalizationToolDefinition.exportPenaltyToolOLin0802HLin1002(
InputSimulatorDefinition.numberOfAggrItems)
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
dataset: ADataset = DatasetST.readDatasets()
events = dataset.eventsDF
serials = dataset.serialsDF
historyDF: AHistory = HistoryDF("test01")
# Init evalTool
evalTool: AEvalTool = EvalToolContext({
EvalToolContext.ARG_ITEMS:
serials, # ITEMS
EvalToolContext.ARG_EVENTS:
events, # EVENTS (FOR CALCULATING HISTORY OF USER)
EvalToolContext.ARG_DATASET:
"st", # WHAT DATASET ARE WE IN
EvalToolContext.ARG_HISTORY:
historyDF
}) # empty instance of AHistory is OK for ST dataset
pDescr: Portfolio1AggrDescription = Portfolio1AggrDescription(
"ContextDHondtNIF" + jobID, rIDs, rDescs,
InputAggrDefinition.exportADescDContextHondtINF(
selector, pToolOLin0802HLin1002, evalTool))
batchID: str = "stDiv90Ulinear0109R1"
dataset: DatasetST = DatasetST.readDatasets()
behaviourFile: str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF: DataFrame = BehavioursST.readFromFileST(behaviourFile)
model: DataFrame = PModelDHondt(pDescr.getRecommendersIDs())
print(model)
# simulation of portfolio
simulator: Simulator = Simulator(batchID, SimulationST, argsSimulationDict,
dataset, behavioursDF)
simulator.simulate([pDescr], [model], [evalTool],
[HistoryHierDF(pDescr.getPortfolioID())])
def test21():
print("Simulation: ST FuzzyDHondtDirectOptimize")
jobID: str = "Roulette1"
selector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 3})
pProbToolOLin0802HLin1002: APenalization = PenalizationToolDefinition.exportProbPenaltyToolOStat08HLin1002(
InputSimulatorDefinition.numberOfAggrItems)
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr: Portfolio1AggrDescription = Portfolio1AggrDescription(
"FuzzyDHondtDirectOptimize" + jobID, rIDs, rDescs,
InputAggrDefinition.exportADescDHondtDirectOptimizeThompsonSamplingINF(
selector, pProbToolOLin0802HLin1002, "DCG"))
batchID: str = "stDiv90Ulinear0109R1"
dataset: DatasetST = DatasetST.readDatasets()
behaviourFile: str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF: DataFrame = BehavioursST.readFromFileST(behaviourFile)
model: DataFrame = PModelDHondtBanditsVotes(pDescr.getRecommendersIDs())
#lrClick:float = 0.03
#lrView:float = lrClick / 500
#eTool:AEvalTool = EvalToolDHondt({EvalToolDHondt.ARG_LEARNING_RATE_CLICKS: lrClick,
# EvalToolDHondt.ARG_LEARNING_RATE_VIEWS: lrView})
eTool: AEvalTool = EvalToolDHondtBanditVotes({})
# simulation of portfolio
simulator: Simulator = Simulator(batchID, SimulationST, argsSimulationDict,
dataset, behavioursDF)
simulator.simulate([pDescr], [model], [eTool],
[HistoryHierDF(pDescr.getPortfolioID())])
def getFuzzyDHontThompsonSamplingINF():
taskID:str = "Web" + "FuzzyDHondtThompsonSamplingINF" + "Fixed" + "OLin0802HLin1002"
selector:ADHondtSelector = TheMostVotedItemSelector({})
penalization:APenalization = PenalizationToolDefinition.exportProbPenaltyToolOLin0802HLin1002(20)
aDescDHont:AggregationDescription = InputAggrDefinition.exportADescDHondtThompsonSamplingINF(selector, penalization)
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr:Portfolio1AggrDescription = Portfolio1AggrDescription(
taskID, rIDs, rDescs, aDescDHont)
history:AHistory = HistoryHierDF(taskID)
dataset:ADataset = DatasetST.readDatasets()
port:APortfolio = pDescr.exportPortfolio(taskID, history)
port.train(history, dataset)
model:DataFrame = PModelDHondtBanditsVotes(pDescr.getRecommendersIDs())
evalTool:AEvalTool = EvalToolDHondtBanditVotes({})
return (taskID, port, model, evalTool, history)
def getFuzzyDHontINF():
#taskID:str = "Web" + "FuzzyDHondtINF" + "Roulette1"
taskID:str = "Web" + "FuzzyDHondt" + "Fixed"
dataset:ADataset = DatasetST.readDatasets()
#selector:ADHondtSelector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
selector:ADHondtSelector = TheMostVotedItemSelector({})
pToolOLin0802HLin1002:APenalization = PenalizationToolDefinition.exportPenaltyToolOLin0802HLin1002(
InputSimulatorDefinition.numberOfAggrItems)
aDescDHont:AggregationDescription = InputAggrDefinition.exportADescDHondtINF(selector, pToolOLin0802HLin1002)
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr:Portfolio1AggrDescription = Portfolio1AggrDescription(
taskID, rIDs, rDescs, aDescDHont)
history:AHistory = HistoryHierDF(taskID)
port:APortfolio = pDescr.exportPortfolio(taskID, history)
port.train(history, dataset)
model:DataFrame = PModelDHondt(pDescr.getRecommendersIDs())
evalTool:AEvalTool = EvalToolDHondt({EvalToolDHondt.ARG_LEARNING_RATE_CLICKS: 0.03,
EvalToolDHondt.ARG_LEARNING_RATE_VIEWS: 0.03 / 500})
return (taskID, port, model, evalTool, history)
def getContextFuzzyDHondtDirectOptimizeINF():
# taskID:str = "Web" + "ContextFuzzyDHondtDirectOptimizeINF" + "Roulette1"
taskID:str = "Web" + "ContextFuzzyDHondtDirectOptimizeINF" + "Fixed"
dataset:ADataset = DatasetST.readDatasets()
# selector:ADHondtSelector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
selector:ADHondtSelector = TheMostVotedItemSelector({})
pToolOLin0802HLin1002:APenalization = PenalizationToolDefinition.exportProbPenaltyToolOLin0802HLin1002(
InputSimulatorDefinition.numberOfAggrItems)
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
history:AHistory = HistoryHierDF(taskID)
# Init eTool
evalTool:AEvalTool = EvalToolContext({
EvalToolContext.ARG_ITEMS: dataset.serialsDF, # ITEMS
EvalToolContext.ARG_EVENTS: dataset.eventsDF, # EVENTS (FOR CALCULATING HISTORY OF USER)
EvalToolContext.ARG_DATASET: "st", # WHAT DATASET ARE WE IN
EvalToolContext.ARG_HISTORY: history}) # empty instance of AHistory is OK for ST dataset
aDescDHont:AggregationDescription = InputAggrDefinition.exportADescDContextHondtDirectOptimizeINF(selector, pToolOLin0802HLin1002, evalTool)
pDescr:Portfolio1AggrDescription = Portfolio1AggrDescription(
taskID, rIDs, rDescs, aDescDHont)
port:APortfolio = pDescr.exportPortfolio(taskID, history)
port.train(history, dataset)
model:DataFrame = PModelDHondt(pDescr.getRecommendersIDs())
return (taskID, port, model, evalTool, history)
def test03():
print("Test 03")
print("Running Recommender BPRMF on ST:")
batchID: str = "batchID"
dataset: ADataset = DatasetST.readDatasets()
history: AHistory = HistoryHierDF(["aa"])
argumentsDict: Dict[str, object] = {
RecommenderBPRMF.ARG_EPOCHS: 2,
RecommenderBPRMF.ARG_FACTORS: 10,
RecommenderBPRMF.ARG_LEARNINGRATE: 0.05,
RecommenderBPRMF.ARG_UREGULARIZATION: 0.0025,
RecommenderBPRMF.ARG_BREGULARIZATION: 0,
RecommenderBPRMF.ARG_PIREGULARIZATION: 0.0025,
RecommenderBPRMF.ARG_NIREGULARIZATION: 0.00025
}
r: ARecommender = RecommenderBPRMF(batchID, argumentsDict)
r.train(history, dataset)
numberOfItems: int = 20
userId: int = 62302
res = r.recommend(userId, numberOfItems, argumentsDict)
print(res)
userId: int = 3462303
res = r.recommend(userId, numberOfItems, argumentsDict)
print(res)
def test21():
print("Simulation: ST DHontThompsonSampling")
jobID: str = "Roulette1"
selector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr: Portfolio1AggrDescription = Portfolio1AggrDescription(
"DHontThompsonSampling" + jobID, rIDs, rDescs,
InputAggrDefinition.exportADescDHondtThompsonSampling(selector))
batchID: str = "stDiv90Ulinear0109R1"
dataset: DatasetST = DatasetST.readDatasets()
behaviourFile: str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF: DataFrame = BehavioursST.readFromFileST(behaviourFile)
model: DataFrame = PModelDHondtBanditsVotes(pDescr.getRecommendersIDs())
print(model)
evalTool: AEvalTool = EvalToolDHondtBanditVotes({})
# simulation of portfolio
simulator: Simulator = Simulator(batchID, SimulationST, argsSimulationDict,
dataset, behavioursDF)
simulator.simulate([pDescr], [model], [evalTool],
[HistoryHierDF(pDescr.getPortfolioID())])
def exportSimulatorSlantour(self, batchID: str,
divisionDatasetPercentualSize: int,
uBehaviourID: str, repetition: int):
argsSimulationDict: dict = {
SimulationML.ARG_WINDOW_SIZE: 5,
SimulationML.ARG_RECOM_REPETITION_COUNT: repetition,
SimulationML.ARG_NUMBER_OF_RECOMM_ITEMS: 100,
SimulationML.ARG_NUMBER_OF_AGGR_ITEMS: self.numberOfAggrItems,
SimulationML.ARG_DIV_DATASET_PERC_SIZE:
divisionDatasetPercentualSize,
SimulationML.ARG_HISTORY_LENGTH: 10
}
# dataset reading
dataset: ADataset = DatasetST.readDatasets()
behaviourFile: str = BehavioursST.getFile(uBehaviourID)
behavioursDF: DataFrame = BehavioursST.readFromFileST(behaviourFile)
# simulation of portfolio
simulator: Simulator = Simulator(batchID, SimulationST,
argsSimulationDict, dataset,
behavioursDF)
return simulator
def test03():
print("Test 03")
print("Running RecommenderW2V ST:")
dataset: DatasetST = DatasetST.readDatasets()
trainDataset: DatasetST = dataset
eventsDF: DataFrame = dataset.eventsDF
# train recommender
argsDict: Dict[str, str] = {
RecommenderW2V.ARG_ITERATIONS: 50000,
RecommenderW2V.ARG_TRAIN_VARIANT: 'all',
RecommenderW2V.ARG_USER_PROFILE_SIZE: 5,
RecommenderW2V.ARG_USER_PROFILE_STRATEGY: 'weightedMean',
RecommenderW2V.ARG_VECTOR_SIZE: 32,
RecommenderW2V.ARG_WINDOW_SIZE: 1,
RecommenderW2V.ARG_ALLOWED_ITEMIDS: list(range(0, 1000))
}
rec: ARecommender = RecommenderW2V("RecommenderW2V", argsDict)
rec.train(HistoryDF("test03"), trainDataset)
uDF: DataFrame = DataFrame([eventsDF.iloc[9000]])
print(uDF)
rec.update(uDF, argsDict)
r: Series = rec.recommend(2760420, 50, argsDict)
print(r)
def test01():
print("Test 01")
rDescr: RecommenderDescription = RecommenderDescription(
RecommenderTheMostPopular, {})
recommenderID: str = "TheMostPopular"
pDescr: Portfolio1MethDescription = Portfolio1MethDescription(
recommenderID.title(), recommenderID, rDescr)
dataset: ADataset = DatasetST.readDatasets()
history: AHistory = HistoryDF("test")
p: APortfolio = pDescr.exportPortfolio("jobID", history)
portFolioModel: DataFrame = DataFrame()
p.train(history, dataset)
df: DataFrame = DataFrame(
[[1, 555]], columns=[Events.COL_USER_ID, Events.COL_OBJECT_ID])
p.update(ARecommender.UPDT_CLICK, df)
userID: int = 1
r, rp = p.recommend(userID, portFolioModel,
{APortfolio.ARG_NUMBER_OF_AGGR_ITEMS: 20})
print(r)
def test04():
print("Test 04")
print("Running Recommender BPRMF on SL:")
from datasets.slantour.events import Events # class
dataset: DatasetST = DatasetST.readDatasets()
trainDataset: DatasetST = dataset
eventsDF: DataFrame = dataset.eventsDF
uIDMax: int = eventsDF[Events.COL_USER_ID].max()
print("uIDMax: " + str(uIDMax))
iIDMax: int = eventsDF[Events.COL_OBJECT_ID].max()
print("iIDMax: " + str(iIDMax))
# train recommender
rec: ARecommender = RecommenderBPRMFImplicit(
"test", {
RecommenderBPRMFImplicit.ARG_FACTORS: 20,
RecommenderBPRMFImplicit.ARG_ITERATIONS: 50,
RecommenderBPRMFImplicit.ARG_LEARNINGRATE: 0.003,
RecommenderBPRMFImplicit.ARG_REGULARIZATION: 0.003
})
rec.train(HistoryDF("test04"), trainDataset)
uDF1: DataFrame = DataFrame([eventsDF.iloc[9000]])
print(uDF1)
rec.update(uDF1, {})
userID1: int = uIDMax + 1
itemID1: int = iIDMax + 1
itemID2: int = iIDMax + 2
# update with unknown user and unknown item
uDF2: DataFrame = DataFrame(
columns=[Events.COL_USER_ID, Events.COL_OBJECT_ID])
uDF2.loc[0] = [userID1, itemID1]
print(uDF2)
rec.update(uDF2, {})
# update with unknown item
uDF3: DataFrame = DataFrame(
columns=[Events.COL_USER_ID, Events.COL_OBJECT_ID])
uDF3.loc[0] = [userID1, itemID2]
print(uDF3)
rec.update(uDF3, {})
r: Series = rec.recommend(23, 50, {})
print(r)
print()
# recommend for unknown user
r: Series = rec.recommend(userID1, 50, {})
print(r)
def test01():
print("Test 01")
dataset: ADataset = DatasetST.readDatasets()
print(dataset.serialsDF.columns)
print(dataset.serialsDF.head(10))
print(dataset.eventsDF.columns)
print(dataset.eventsDF.head(10))
def test03():
print("Test 03")
print("Running RecommenderCosineCB ST:")
dataset: DatasetST = DatasetST.readDatasets()
args: dict = {
RecommenderCosineCB.ARG_CB_DATA_PATH:
Configuration.cbSTDataFileWithPathTFIDF,
# RecommenderCosineCB.ARG_CB_DATA_PATH:Configuration.cbSTDataFileWithPathOHE,
RecommenderCosineCB.ARG_USER_PROFILE_SIZE:
5,
RecommenderCosineCB.ARG_USER_PROFILE_STRATEGY:
"max",
RecommenderCosineCB.ARG_USE_DIVERSITY:
False,
RecommenderCosineCB.ARG_ALLOWED_ITEMIDS:
list(range(0, 1000))
} #True
rec: ARecommender = RecommenderCosineCB("test", args)
rec.train(HistoryDF("test"), dataset)
#eventsDFDFUpdate:DataFrame = dataset.eventsDF.iloc[5003:5004]
#print(eventsDFDFUpdate)
#rec.update(eventsDFDFUpdate, args)
# user with very outdated profile - no recent objects
r: Series = rec.recommend(3500678, 10, args)
#print(type(r))
print(r)
args: dict = {
RecommenderCosineCB.ARG_CB_DATA_PATH:
Configuration.cbSTDataFileWithPathTFIDF,
# RecommenderCosineCB.ARG_CB_DATA_PATH:Configuration.cbSTDataFileWithPathOHE,
RecommenderCosineCB.ARG_USER_PROFILE_SIZE:
5,
RecommenderCosineCB.ARG_USER_PROFILE_STRATEGY:
"max",
RecommenderCosineCB.ARG_USE_DIVERSITY:
True,
RecommenderCosineCB.ARG_MMR_LAMBDA:
0.5,
RecommenderCosineCB.ARG_ALLOWED_ITEMIDS:
list(range(0, 1000))
} #True
rec: ARecommender = RecommenderCosineCB("test", args)
rec.train(HistoryDF("test"), dataset)
r: Series = rec.recommend(3500678, 10, args)
print(r)
"""
def test03():
print("Test 03")
dataset: ADataset = DatasetST.readDatasets()
print(dataset.getTheMostSold())
#eventsDF:DataFrame = dataset.eventsDF
#eventsDF = eventsDF.loc[eventsDF[Events.COL_OBJECT_ID] != 0]
print(dataset.eventsDF.columns)
print(dataset.serialsDF.columns)
print(dataset.eventsDF["userID"].max())
def test03():
print("Test 03")
print("Running RecommenderVSKNN ST:")
dataset: DatasetST = DatasetST.readDatasets()
testD = dataset.eventsDF.iloc[4000:10000]
dataset.eventsDF = dataset.eventsDF.iloc[0:4000]
# train recommender
rec: ARecommender = RecommenderVMContextKNN("test", {})
start = time.time()
rec.train(HistoryDF("test03"), dataset)
end = time.time()
print("Time to train: " + str(end - start))
i = 0
"""
maxI = 5990
while i < maxI:
eventsDFDFUpdate:DataFrame = dataset.eventsDF.iloc[i:i+1]
rec.update(rec.UPDT_CLICK, eventsDFDFUpdate)
i = i+1
if i%100 == 0:
print(i)
"""
r: Series = rec.recommend(3342336, 20,
{rec.ARG_ALLOWED_ITEMIDS: list(range(0, 1000))})
print(type(r))
print(r)
r: Series = rec.recommend(2035310, 20,
{rec.ARG_ALLOWED_ITEMIDS: list(range(0, 1000))})
print(type(r))
print(r)
r: Series = rec.recommend(3342341, 20,
{rec.ARG_ALLOWED_ITEMIDS: list(range(0, 1000))})
print(type(r))
print(r)
# testing of a non-existent user
r: Series = rec.recommend(10000, 50,
{rec.ARG_ALLOWED_ITEMIDS: list(range(0, 1000))})
print(type(r))
print(r)
print("================== END OF TEST 03 ======================\n\n\n\n\n")
def run(self, batchID: str, jobID: str):
divisionDatasetPercentualSize: int
uBehaviour: str
repetition: int
divisionDatasetPercentualSize, uBehaviour, repetition = InputABatchDefinition(
).getBatchParameters(self.datasetID)[batchID]
selector: ADHondtSelector
selector, negativeImplFeedback = self.getParameters()[jobID]
portfolioID: str = self.getBatchName() + jobID
history: AHistory = HistoryHierDF(portfolioID)
dataset: ADataset = DatasetST.readDatasets()
events = dataset.eventsDF
serials = dataset.serialsDF
historyDF: AHistory = HistoryHierDF("test01")
# Init evalTool
evalTool: AEvalTool = EvalToolContext({
EvalToolContext.ARG_ITEMS:
serials, # ITEMS
EvalToolContext.ARG_EVENTS:
events, # EVENTS (FOR CALCULATING HISTORY OF USER)
EvalToolContext.ARG_DATASET:
"st", # WHAT DATASET ARE WE IN
EvalToolContext.ARG_HISTORY:
historyDF
}) # empty instance of AHistory is OK for ST dataset
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs(
)
aDescContextDHont: AggregationDescription = InputAggrDefinition.exportADescDContextHondtINF(
selector, negativeImplFeedback, evalTool)
pDescr: Portfolio1AggrDescription = Portfolio1AggrDescription(
portfolioID, rIDs, rDescs, aDescContextDHont)
model: DataFrame = PModelDHondt(pDescr.getRecommendersIDs())
simulator: Simulator = InputSimulatorDefinition(
).exportSimulatorSlantour(batchID, divisionDatasetPercentualSize,
uBehaviour, repetition)
simulator.simulate([pDescr], [model], [evalTool], [history])
def test24():
print("Simulation: ST CB")
rDescr:RecommenderDescription = InputRecomMLDefinition.exportRDescCosineCBcbdOHEupsweightedMeanups3()
pDescr:APortfolioDescription = Portfolio1MethDescription(InputRecomMLDefinition.COS_CB_MEAN.title(),
InputRecomMLDefinition.COS_CB_MEAN, rDescr)
batchID:str = "slantourDiv90Ulinear0109R1"
dataset:DatasetST = DatasetST.readDatasets()
behaviourFile:str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF:DataFrame = BehavioursST.readFromFileST(behaviourFile)
# simulation of portfolio
simulator:Simulator = Simulator(batchID, SimulationST, argsSimulationDict, dataset, behavioursDF)
simulator.simulate([pDescr], [DataFrame()], [EToolDoNothing({})], [HistoryHierDF(pDescr.getPortfolioID())])
def test26():
print("Simulation: ST VMCMF")
rDescr:RecommenderDescription = InputRecomSTDefinition.exportRDescVMContextKNN()
pDescr:APortfolioDescription = Portfolio1MethDescription(InputRecomSTDefinition.VMC_KNN.title(),
InputRecomSTDefinition.VMC_KNN, rDescr)
batchID:str = "slantourDiv90Ulinear0109R1"
dataset:DatasetST = DatasetST.readDatasets()
behaviourFile:str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF:DataFrame = BehavioursST.readFromFileST(behaviourFile)
# simulation of portfolio
simulator:Simulator = Simulator(batchID, SimulationST, argsSimulationDict, dataset, behavioursDF)
simulator.simulate([pDescr], [DataFrame()], [EToolDoNothing({})], [HistoryHierDF(pDescr.getPortfolioID())])
def test02():
print("Test 02")
dataset: ADataset = DatasetST.readDatasets()
pd.set_option('display.max_rows', None)
pd.set_option('display.max_columns', None)
pd.set_option('display.width', None)
pd.set_option('display.max_colwidth', -1)
eventsDF: DataFrame = dataset.eventsDF
eventsDF = eventsDF.loc[eventsDF[Events.COL_OBJECT_ID] != 0]
print(eventsDF.columns)
print(eventsDF[[
Events.COL_VISIT_ID, Events.COL_USER_ID, Events.COL_OBJECT_ID
]].head(100))
def test13():
print("Test 13")
print("Running Recommender BPRMF on ST:")
from datasets.slantour.events import Events # class
batchID: str = "batchID"
trainDataset: ADataset
testDataset: ADataset
trainDataset, testDataset = DatasetST.readDatasets().divideDataset(90)
testUserIDs: ndarray = testDataset.eventsDF[Events.COL_USER_ID].unique()
history: AHistory = HistoryHierDF(["aa"])
numberOfItems: int = 20
rd: RecommenderDescription = InputRecomSTDefinition.exportRDescBPRMFIMPLf50i20lr01r003(
)
rd: RecommenderDescription = InputRecomSTDefinition.exportRDescBPRMF()
#rd:RecommenderDescription = InputRecomSTDefinition.exportRDescTheMostPopular()
#rd:RecommenderDescription = InputRecomSTDefinition.exportRDescKNN()
#rd:RecommenderDescription = InputRecomSTDefinition.exportRDescCosineCBcbdOHEupsweightedMeanups5()
#rd:RecommenderDescription = InputRecomSTDefinition.exportRDescW2Vtalli100000ws1vs32upsmaxups1()
#rd:RecommenderDescription = InputRecomSTDefinition.exportRDescW2talli200000ws1vs64upsweightedMeanups5()
r: ARecommender = rd.exportRecommender("aaa")
argumentsDict: Dict = rd.getArguments()
r.train(history, trainDataset)
numberOfHit: int = 0
for userIdI in testUserIDs[0:1000]:
recI: Series = r.recommend(int(userIdI), numberOfItems, argumentsDict)
recItemIDsI: List[int] = [i for i in recI.keys()]
windowItemIds: List[int] = testDataset.eventsDF.loc[
testDataset.eventsDF[Events.COL_USER_ID] == userIdI][
Events.COL_OBJECT_ID].unique()
itemIdsHitted: List[int] = list(set(recItemIDsI) & set(windowItemIds))
numberOfHit += len(itemIdsHitted)
print("")
print("numberOfHit: " + str(numberOfHit))
def getW2Vtalli100000ws1vs32upsmaxups1():
taskID:str = "Web" + "W2Vtalli100000ws1vs32upsmaxups1"
rDescr:RecommenderDescription = InputRecomSTDefinition.exportRDescW2Vtalli100000ws1vs32upsmaxups1()
recommenderID:str = "W2V"
pDescr:Portfolio1MethDescription = Portfolio1MethDescription(recommenderID.title(), recommenderID, rDescr)
dataset:ADataset = DatasetST.readDatasets()
history:AHistory = HistoryHierDF(taskID)
port:APortfolio = pDescr.exportPortfolio(taskID, history)
port.train(history, dataset)
model:DataFrame = DataFrame()
evalTool:AEvalTool = EToolDoNothing({})
return (taskID, port, model, evalTool, history)
def getBanditTS():
taskID:str = "Web" + "BanditTS" + "Roulette1"
selector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr:Portfolio1AggrDescription = Portfolio1AggrDescription(
taskID, rIDs, rDescs, InputAggrDefinition.exportADescBanditTS(selector))
dataset:DatasetST = DatasetST.readDatasets()
history:AHistory = HistoryHierDF(taskID)
port:APortfolio = pDescr.exportPortfolio(taskID, history)
port.train(history, dataset)
model:DataFrame = PModelBandit(pDescr.getRecommendersIDs())
evalTool:AEvalTool = EvalToolBanditTS({})
return (taskID, port, model, evalTool, history)
def visualizationST():
print("visualizationST")
from datasets.slantour.events import Events # class
# dataset reading
dataset: ADataset = DatasetST.readDatasets()
#dataset:ADataset = DatasetST.readDatasetsSkipOutlierUsers(500)
print(dataset.eventsDF.head())
userIdsWithDuplicites: List[int] = dataset.eventsDF[
Events.COL_USER_ID].tolist()
userIds: List[int] = list(set(userIdsWithDuplicites))
plt.hist(userIdsWithDuplicites,
len(userIds),
None,
fc='none',
lw=1.5,
histtype='step')
plt.ticklabel_format(style='plain')
plt.show()
def visualizationST2():
print("visualizationST2")
from datasets.slantour.events import Events # class
dataset: ADataset = DatasetST.readDatasets()
eventsDF: Events = dataset.eventsDF
eventsDF.sort_values(by=[Events.COL_START_DATE_TIME], inplace=True)
userIdsWithDuplicites: List[int] = eventsDF[Events.COL_USER_ID].tolist()
userIds: List[int] = list(set(userIdsWithDuplicites))
print("userIdsCount: " + str(len(userIds)))
#return
for userIdI in userIds:
indexesOfUserI: List[int] = eventsDF[eventsDF[Events.COL_USER_ID] ==
userIdI].index.tolist()
print(indexesOfUserI)
distanceI: int = max(indexesOfUserI) - min(indexesOfUserI)
print(distanceI)
def test21():
print("Simulation: ST FuzzyDHondt")
jobID: str = "Roulette1"
selector = RouletteWheelSelector({RouletteWheelSelector.ARG_EXPONENT: 1})
rIDs, rDescs = InputRecomSTDefinition.exportPairOfRecomIdsAndRecomDescrs()
pDescr: Portfolio1AggrDescription = Portfolio1AggrDescription(
"FuzzyDHondt" + jobID, rIDs, rDescs,
InputAggrDefinition.exportADescDHondt(selector))
batchID: str = "stDiv90Ulinear0109R1"
dataset: DatasetST = DatasetST.readDatasets()
behaviourFile: str = BehavioursST.getFile(BehavioursST.BHVR_LINEAR0109)
behavioursDF: DataFrame = BehavioursST.readFromFileST(behaviourFile)
model: DataFrame = PModelDHondt(pDescr.getRecommendersIDs())
print(model)
lrClick: float = 0.1
lrView: float = lrClick / 300
evalTool: AEvalTool = EvalToolDHondt({
EvalToolDHondt.ARG_LEARNING_RATE_CLICKS:
lrClick,
EvalToolDHondt.ARG_LEARNING_RATE_VIEWS:
lrView
})
# simulation of portfolio
simulator: Simulator = Simulator(batchID, SimulationST, argsSimulationDict,
dataset, behavioursDF)
simulator.simulate([pDescr], [model], [evalTool],
[HistoryHierDF(pDescr.getPortfolioID())])
def test03():
# First get Dataset Data
dataset: ADataset = DatasetST.readDatasets()
events = dataset.eventsDF
serials = dataset.serialsDF
# I created some dummy data, but each key,value pair should be result list from a recommender
# (=what recommender recommended)
methodsResultDict: dict = {
"metoda1":
pd.Series([0.2, 0.1, 0.3, 0.3, 0.1], [32, 2, 8, 1, 4], name="rating"),
"metoda2":
pd.Series([0.1, 0.1, 0.2, 0.3, 0.3], [1, 5, 32, 6, 7], name="rating")
}
# init votes for each recommender
portfolioModelData = [['metoda1', 0.6], ['metoda2', 0.4]]
portfolioModelDF: DataFrame = pd.DataFrame(portfolioModelData,
columns=["methodID", "votes"])
portfolioModelDF.set_index("methodID", inplace=True)
userID = 1
itemID = 20
historyDF: AHistory = HistoryDF("test01")
# WHAT EVALUATIOR NEEDS into dictionary!
evaluationDict: dict = {
EvalToolContext.ARG_USER_ID: userID,
EvalToolContext.ARG_ITEM_ID:
itemID, # ITEMID (not mandatory if EvalToolContext.ARG_PAGE_TYPE != "zobrazit")
EvalToolContext.ARG_SENIORITY: 5, # SENIORITY OF USER
EvalToolContext.ARG_PAGE_TYPE:
"zobrazit", # TYPE OF PAGE ("zobrazit", "index" or "katalog)
EvalToolContext.ARG_ITEMS_SHOWN: 10 # HOW MANY ITEMS ARE SHOWN TO USER
}
# Init eTool
eToolContext = EvalToolContext({
EvalToolContext.ARG_ITEMS: serials, # ITEMS
EvalToolContext.ARG_EVENTS:
events, # EVENTS (FOR CALCULATING HISTORY OF USER)
EvalToolContext.ARG_DATASET: "st", # WHAT DATASET ARE WE IN
EvalToolContext.ARG_HISTORY: historyDF
} # empty instance of AHistory is OK for ST dataset
)
aggr: AggrContextFuzzyDHondt = AggrContextFuzzyDHondt(
historyDF,
{ # empty instance of AHistory is OK for ST dataset
AggrContextFuzzyDHondt.ARG_EVAL_TOOL: eToolContext, # eTool
AggrContextFuzzyDHondt.ARG_SELECTOR: TheMostVotedItemSelector(
{}) # ? FuzzyDHondt needs this, not contextAggr
})
# Get data from aggregator
rItemsWithResponsibility = aggr.runWithResponsibility(
methodsResultDict,
portfolioModelDF,
userID,
numberOfItems=5,
argumentsDict=evaluationDict)
# call click & displayed methods
l1 = eToolContext.displayed(rItemsWithResponsibility, portfolioModelDF,
evaluationDict)
# rItemsWithResponsibility[0][0] is clicked item
print(rItemsWithResponsibility)
l1 = eToolContext.click(rItemsWithResponsibility,
rItemsWithResponsibility[0][0], portfolioModelDF,
evaluationDict)
# ...
# ...
# ...
# user is now on "index" page type, so we have to change page type in evaluationDict (!)
evaluationDict[EvalToolContext.ARG_PAGE_TYPE] = "index"
# same as before
# Get data from aggregator
rItemsWithResponsibility = aggr.runWithResponsibility(
methodsResultDict,
portfolioModelDF,
userID,
numberOfItems=5,
argumentsDict=evaluationDict)
# call click & displayed methods
l1 = eToolContext.displayed(rItemsWithResponsibility, portfolioModelDF,
evaluationDict)
# rItemsWithResponsibility[0][0] is clicked item
l1 = eToolContext.click(rItemsWithResponsibility,
rItemsWithResponsibility[0][0], portfolioModelDF,
evaluationDict)