On daily-sampled series, try running weekly forecasts, then upscaling

[capelastegui]

We have found that, in some cases, for the same time series, a weekly sampled forecast may perform better than a daily sampled one. We could try, for daily sampled series, running weekly sampled forecasts and re-upscaling, and comparing it with the regular daily sampled forecasts.

[bezes]

I concur. I would also generalise the solution to other frequencies, too, e.g. convert from daily to monthly, etc.

[capelastegui]

This is now easier due to #198