Possibility of reverse-midas use with midas_r_ic_table and expand_weight_lags?

[elemn]

Hello,

First off, thank you for writing this quick guide to running reverse-MIDAS models, it's extremely helpful.

I am wondering if it is possible to calculate a reverse model which chooses a best variable+lag combination to reduce a common selection criterion such as AIC or BIC. MIDAS documentation points towards midas_r_ic_table and expand_weight_lags, however it's not clear if these support transformations of the dependent variable.

Thank you for your thoughts if you have a chance! I'm new to working with this package and model type and finding it all a bit confusing.