Hi @vagudets
These are some notes I don't want to forget for this development.
When running prepare_outputs() or the functions reading a recipe, we could include eval.exp = TRUE .
recipe <- read_yaml(recipe_file, eval.exp = TRUE)
In this way, if the recipe includes R code it will be executed. In this case, for decadal purposes, a user could include this line in the recipe to aggregate January to March period along the forecast length:
- !expr sort(c(seq(1,120, 12), seq(2,120,13), seq(3, 120, 14)))The recipe could include this section (note that only one method can be use (user_def or ini/end), if both are requested, the code should only use one of the options (default ini/end):
Workflow:
Time_aggregation:
execute: yes #do we need this?
user_def:
- [1, 3] # aggregate from 1 to 3 forecast times
- !expr sort(c(seq(1,120, 12), seq(2,120,13), seq(3, 120, 14))) # aggregate 1,2,3,13,14,15,...
method: average #accumulated
ini: [1, 2, 1] # aggregate from 1 to 2; from 2 to 3 and from 1 to 3
end: [2, 3, 3]Cheers,
Núria