A moving average that weighted with Farey fractions. It matches a standard linear weighted average almost one-to-one. Why? Because both averages have strictly monotonic weighting sequences and assign a higher weight to latests data. So, Farey weights are just scaled to [0, 1] linear ones. Instead of specifing period you specify an order of Farey sequence. To learn more about Farey sequence you can refer to Wiki
Published just for reference, it is not intended for trading purposes.