r_hat >= 1.1 when using offline channels with specific pace

[TomasPapantos]

Hi!

I've ran the LMMM with the online channels and the model worked well, but when I added the offline channels (which don't have data on all weeks - so 80%+ of rows are 0), when running the forecasting portion of LMMM the r_hat is >= 1.1 (extremely high values).

I've tried adding all of those up into a single "offline" channel but it didn't work.

Does anyone know hoy to overcome this?

Thanks in advanced.

Tom