Univariate GARCH-MIDAS, Double-Asymmetric GARCH-MIDAS and MEM-MIDAS

Adds the MIxing Data Sampling (MIDAS, Ghysels et al. (2007) ) components within the GARCH (Engle et al. (2013) ) and MEM (Engle (2002) ) frameworks, with the aim of predicting the volatility with additional low-frequency (that is, MIDAS) terms. The estimation takes place through simple functions, which provide in-sample and (if present) and out-of-sample evaluations. 'rumidas' also offers a summary tool, which synthesizes the main information of the estimated model. Finally, an option to generate one-step-ahead volatility forecasts automatically divides the whole period into a training and testing samples.