Hierarchical time series forecasting optimal reconciliation using Fable in R

Question

I am doing hierarchical time series forecasting using fable package in R. I am using the optimal reconciliation method to reconcile the forecast. Here is the example code.

agg_sw <- df %>%
  aggregate_key(productcategory/brand/sku, sales = sum(sales))

# Fit the model
ets_fit <- agg_sw %>%
  model(ets = ETS(sales)) %>%
  reconcile(ols = min_trace(ets, method = "ols"))

# Forecast
fc <- forecast(ets_fit, h= "1 year")

Is it possible to use a different forecasting method at each level (e.g., sku/brand/product) and reconcile? If so, kindly let me know how to do it.

Answer 1

You can use any model that you like to forecast each series in the hierarchy. However, the fable package only supports a few forecasting models including ets, tslm for automatic reconciliations. If you want to use a random forest or deep learning model, you can do so but you need to either write your own wrapper to use fable (see https://fabletools.tidyverts.org/articles/extension_models.html), or write the reconciliation function separately to do the hierarchical forecasting reconciliations.