Hi @vagudets,
This development is to include the significance of the metrics mean-bias and spread-to-error ratio. For this the function Bias.R in the skill module has been updated, and a new function SprErr.R has been included. These functions are currently called from the folder Skill/R/temp/, since there are not yet included in the master branch, see below the two issues regarding these developments:
A new statistic spread, has also been included in the Statistics module, to be able to calculate the spread-to-error ratio significance in the scorecards.
Additionally, the scorecards code has been restructured. The code is now separated into two functions; scorecard_calculations.R and scorecards_plotting.R.
The function scorecard_calculations applied to the atomic recipes and calculates the spatially aggregated metrics and statistical significance to visualise in the scorecards. The function uses as an input skill_metrics and statistics (depending on the requested metrics) and saves the aggregated metrics in a netcdf file per start date.
The function scorecards_plotting applies to the general recipe and loads the aggregated metrics from the saved netcdf files and plots the scorecard visualisations.
Let me know if anything is unclear.
Cheers,
Nadia
FYI @nperez