Reformulated is_call_to_function to use the vault information in our advantage. Now we need to pass the current routine as an argument to calls_to_functions.

Bugfixes to fix issues that appeared when including the sea-ice. Changes include: - Fix in calls_to_functions, dimension of contents, has_operations and calls_to_procedure.

Added a new command line "--additional-source" argument which can be used to apply the changes that are necessary to the preprocessed files also to a different source directory. It will allow to automatically transfer last minute fixes directly to the original source.

Fix in plot_result.py and temporary change in PrecisionAnalysis.py to skip making the result explorer.

Must get the accuracy test out of AutoRPE and provide it as an argument.

Must get the accuracy test out of AutoRPE and provide it as an argument.

Removed old way to manage exceptions, added new exception definitions and the exception solver(even it does not properly work)

Code formatting and skipping create_plots in case it already exists. The function now accepts a force parameter to change this behaviour. Renamed Queue object and added few experiment control methods in the object itself.

Adding result_explorer.py, which creates a html page to better visualize the results of the analysis.

Few things: -Added a method to a Job class to return the variable objects that are used in the Job. -Added the variable summary in the created plots. -Deactivated the advanced refinement for when merging groups. Should debug the problem here.

Few things: -Added a method to a Job class to return the variable objects that are used in the Job. -Added the variable summary in the created plots. -Deactivated the advanced refinement for when merging groups. Should debug the problem here.

Using a red margin for outputs that are outside the limits to help make results better understandable. Probably need a small refactor to keep it as a standalone module.

Change to speed-up function get_variable_by_id by creating a dictionary to index de variables. Few parts are kept to keep compatibility with old vaults, but should be cleaned up later.

Added list_of_failed_variables function to obtain a list with the outputs that don't pass the accuracy test.

Bugfix in evaluate_fitness_from_dictionaries, which was failing for simulations longer than 5 days (more than one value per output)