|new| Addition of the 'coveralls' badge to the README
|new| Addition of a random-seed functionality to ensure reproducible results if needed
|new| If a User instance doesn't have any appliance, its repr method prints the user name, and number of users with a message to mention no appliances are assigned to the instance yet. Before it raised a ValueError.
|new| Documentation template was changed to sphinx wagtail theme to improve the navigation through the documentation sections
|new| Add issue templates for issue creation on github
|new| Automatic testing of the jupyter notebooks of the documentation, to make sure the examples are always running through
|new| Adding depreciation warning to back-compatibility Appliance method in User class to let users know they should use the add_appliance method instead.
|new| Introduction of an Appliance parameter to model productive use duty cycles: continuous_use_duty_cycle
|changed| Expanded and revised documentation, with a particular focus on more and clearer usage examples
|changed| Updated requirements for contributing, including pip dependencies specific to developers
|changed| Expanded test coverage
|changed| Updated .py example input files to match the latest formalism of the RAMP code
|changed| Python version was bumped from 3.8 to 3.10
|changed| Improved the way to run the quantitative tests and the instructions to do so
|fixed| Windows compatibility of path to convert .py to .xlsx
|fixed| rand_peak_enlarge is rounded to be at least 1 so that the peak_time_range is never empty
|fixed| Running .xlsx files form the command line
|fixed| Ignore profile of appliances if any of their functional time or randomly allocated time of use are 0
|fixed| Plotting a cloud of profiles from the command line is fixed
|fixed| jupyter notebooks are up to date with the UseCase class
|fixed| UseCase class usage is now documented
|changed| num_profile variable was changed to num_days
|changed| User class get assigned automatically to a default UseCase instance if not provided
|changed| Delete ramp.core.initialize and ramp.core.stochastic_process, move calc_peak_time_range inside UseCase method
|fixed| conversion of .py files into .xlsx is fixed
|fixed| using .py files is now possible in the command line as well as from IDE
|new| tests for example jupyter notebook (smoke test to see if the notebooks run through)
|new| continuous integration setup
|new| first automated tests
|fixed| installation options have been debugged and the documentation updated accordingly
|fixed| automated download of example applications via the download_example functions now includes previously missing .csv files
|hotfix| added option -o to the terminal command line interface to enable the user to provide output path to save ramp results. This option is also accessible to python users using ofname argument of the ramp/ramp_run.py::run_usecase or the ramp/post_process/post_process.py::export_series functions.
|new| added full software documentation
|fixed| refactored the code in order to improve execution time, use of masks were dismissed
|new| added a way to compute ramp profiles for a usecase using parallel processes via the generate_daily_load_profiles_parallel method of the UseCase class (option -p for command line input)
|new| added a way to run a whole year with different input files for each month for seasonality of parameters (only for the command line, type ramp -h in terminal for more help)
|new| added a way to define date ranges for ramp simulation to get the weekdays automatically and avoid always starting on a monday (only for the command line, type ramp -h in terminal for more help)
|new| added a qualitative testing functionality, accessible via test/test_run.py, to check how code changes affect default outputs
|fixed| the way in which the random switch-on time is computed in the stochastic_process has been changed so that it is sampled with uniform probability from a concatenated set of functioning windows, rather than for each window separately (which led to short windows having higher concentration of switch-on events and demand peaks)
|fixed| the default value for the peak_enlarg parameter has been changed from the mistyped value of 0 to the intended value of 0.15
|new| added a paragraph describing the algorithm of RAMP
|new| refactored the code by moving many of the code from stochastic_process.py into the User class in core.py, used function where code was duplicated
|fixed| the user now gets a warning if the allocated window time is shorter than the provided func_time
|new| add a way to run from .xlsx input files, keeping the back compatibility with .py files (there is possibility to convert .py input files to .xslx)
|new| defined a new class in core.py: UseCase which contains a list of User instances
|fixed| variable names are now PEP8 compatible
|new| added input_file_3 as an example of e-cooking loads
|changed| the way in which input files are called in the ramp_run script has been changed to be more explicit and user-friendly
|changed| the readme.md has been updated to describe the purpose of the 3 provided input files, 1: basic electric appliances, 2: DHW, 3: cooking
|changed| the pubs_list.md has been updated with two new publications
|new| created a CHANGELOG.md file to keep track of code changes from now on
|changed| the repository structure has been modified for better clarity, replicating the structure of the sister-project RAMP-mobility
|changed| changed the way in which the probability of coincident switch-on of several identical appliances owned by a single user is computed. Now, it penalisse less the probability of maximum coincidence for off-peak events
|fixed| s_peak value is now set by default to 0.5, rather than 1. This fixes an unwanted behaviour in how the random.gauss function worked