Can someone help me up in setting the historicWindowSize in the AnomalyLikelyHood of the HTM.java since according to the https://www.sciencedirect.com/science/article/pii/S0925231217309864 the W = 8000 and also if we dont remove the historic samples from historicalScores it will keep growing and eventually system would go in out of memory state .
Fixed sized sliding list has been maintained in the python through :
anomalyLikelihood._historicalScores = collections.deque(maxlen=proto.historicWindowSize)