v1.1.0 testing
Hello @allabres @nmilders @jramon
We are planning on releasing a new version of the Verification Suite within the next few weeks. You may find a list of the changes here: #45 (closed)
I want to ask you to please test that the code in the master branch works with your scripts before we release the new version of the Verification Suite. To do this, I recommend that you create a merge request pull the code into a new branch in your forks.
You might need to modify some things:
- The prepare_outputs() function now performs a check on your recipe by default, in order to detect potential errors. If the check gives you any trouble, please report it here. You can disable the check with the parameter
disable_checks
, e.g.:
recipe <- prepare_outputs(recipe_file, disable_checks = T)
- The entry parameters for compute_skill_metrics(), compute_probabilities(), plot_data() and save_data() have changed:
- 'calibrated_data' in save_data() and plot_data() has been deprecated
-
exp
andobs
in compute_skill_metrics() are now one single parameter calleddata
- compute_probabilities() now returns the probabilities for the fcst as well as the hcst.
See #41 (closed) for the full discussion. In summary, if you had a script like the one in https://earth.bsc.es/gitlab/es/auto-s2s/-/snippets/96, you would have to make the following changes:
# Define the path to the recipe
recipe_file <- "modules/Loading/testing_recipes/recipe_system7c3s-tas.yml"
# Prepare the recipe, generate a unique output directory and a logfile
recipe <- prepare_outputs(recipe_file)
# Load datasets
data <- load_datasets(recipe)
# Calibrate datasets
calibrated_data <- calibrate_datasets(recipe, data = data)
# Compute skill metrics
skill_metrics <- compute_skill_metrics(recipe, data = calibrated_data)
# Compute percentiles and probability bins
probabilities <- compute_probabilities(recipe, data = calibrated_data)
# Export all data to netCDF
save_data(recipe, data = calibrated_data, skill_metrics = skill_metrics, probabilities = probabilities)
# Plot the forecast ensemble mean, skill metrics and most likely terciles
plot_data(recipe, data = calibrated_data, skill_metrics = skill_metrics, probabilities = probabilities,
significance = T)
If anything is unclear or you need help changing your scripts, please let me know. I apologize for the hassle, but the new version includes several important bug fixes, so I encourage you to update your forks ASAP.
You can take a look at the new documentation in the wiki.
Any problems you have can be reported in this issue.
Thank you very much,
Victòria