Implements the Dyad Ratios algorithm for estimating latent
variables from time-series survey data. The algorithm estimates a
latent mood dimension (or two dimensions) from a set of issue opinion
series. Supports annual, quarterly, monthly, and daily aggregation
intervals, optional exponential smoothing, and up to two latent
dimensions. Input data can be provided as a data frame or read from
delimited text files. Based on Stimson's 'MCalc' C++ program.
See Stimson (2018)