ha0ye
commented
5 years ago - what is the effective forecast horizon
- check timescale of data
- number of time points
- when do you choose a different forecasting method over AR1 (default best choice a priori)
- source of data (observations vs. model output)
- measurement (count abundance vs. biomass
- length of time series
- how much effort to spend to gather / scrape additional time series (and munge them into usable formats)