A python package implementing Bayesian dynamic linear models for time series data
Bayesian dynamic linear model (DLM) is a power tool for analyzing time series data. It offers both the efficient way (filter using discounting factor without estimating the two variances) and the accurate way (estimation everything via sampling) for inferecing the time series model. Like the usual state space model, Bayesian DLM enpowers the full flexibility, allowing user to build any model including trends, seasonality and holiday or other control variables.