Many scoring or plotting functions compute averages over "locations" for a number of different grouping facets. We say "locations" because this most often the geo_value that gets averaged over, while the groupings are the forecaster, forecast horizon, and forecast date. But under other combinations may be desired.
intersect_averagers(cards, grp_vars, avg_vars)
| cards | long data frame |
|---|---|
| grp_vars | character vector of indicating variables to group on |
| avg_vars | character vector of variables to average over |
a data frame of the same type as input
In the case that we want to make comparisons, we want the avg_vars to be common. This function finds common avg_vars. An example would be if one forecaster makes predictions for a location that others don't, we would want to throw it when ever we compute error measures.