This merge request is to track the changes in this branch. The changes made here are to allow the calculation of the aggregated significance of the skill metrics for the scorecards.
New recipe checks have been included to ensure that the required metrics for the scorecards are computed. This change applies to the script: tools/check_recipe.R.
The following new metrics have been included in the modules/Skill/Skill.R script; crps_clim_syear, crps_syear, rps_clim_syear and rps_syear. These metrics are not temporally aggregated so include an additional dimension: syear. Due to this they cannot be saved with the current function save_metrics. The s2dv functions CRPS and RPS have been updated to include the parameter return_mean to allows the syear dimension to be kept when return_mean = False. These changes are not yet applied in s2dv so can be found in the folder; modules/Skill/R/tmp/.
A new module has been included, modules/Statistics/Statistics.R. Currently this calculates the following statistics; covariance, variance and standard deviation. These statistics have the following dimensions:
> dim(statistics$cov)
dat var sday sweek time latitude longitude
1 1 1 1 2 6 11 Both the new syear metrics and the statistics cannot be saved with the current saving functions, so a new saving function using the CSTool function: CST_SaveExp, will be created to save these metrics and statistics.
FYI @erifarov