I'm using the forecast R package. I have the same dilemma as How to forecast daily time series with weekends and holidays?. I understand that the answers consensus is to model all data within a single time-series.

However, I would like to benchmark the second approach, in which there are 2 separate time-series objects, for workdays and for weekends.

How can I elegantly wrap around these 2 time-series so that I can use the tsCV utility for forecasting 3 months ahead, with each forecast using the correct model (weekend/workday)?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.