Stream Moving Average with Recent-Value Weighting

Given a stream of numeric values and a window size, compute the moving average over the last N items. Follow-ups ask for a weighted moving average that emphasizes recent values and whether it can be done without storing all prior values. Requirements Process an integer or numeric stream online. Given a window size N or X, return the average of the most recent window. Maintain enough state to update the average as new values arrive. Follow-ups include: Recent values should carry higher weights than older values. Avoid storing all historical values. Explain what state is still necessary for exact fixed-window behavior. Notes Candidates explicitly describe this as a coding round for MLE/MLSE; it is not a model-building exercise. Clarify whether the weighted follow-up is an exact finite-window weighted average or an exponentially weighted moving average. The latter has a much smaller state footprint. Preparation Implement a queue plus running sum for fixed-window average. Implement both finite weighted-window and exponential moving average variants, and be ready to explain the memory difference.