GenTS Release v0.8.0

Features and Improvements ✨

• Consolidated all time series manipulations into new TSCollection class
• Finished HFCollection to TSCollection workflow for handling user experience
• Added list/dict functionality to GenTS classes for better user interactivity
• Added user guide to documentation
• Finished API docustrings
• Added per-variable compression and filtering
• Added path modification with swaps (not limited to just directory names)
• Added development environment
• Added serial compatibility (through API)

Bug Fixes 🐛

• Added automatic metadata-pulls for functions that require it
• Removed CLI to avoid errors (may be re-added in the future)
• Removed default multi-primary variable support in main generation code to simplify arguments
• Fixed output date-time file naming error
• Fixed overwrite error when time series and history files appear in the same directory
• Fixed deep copy error that broke some HFCollection functionality

Planned 📋

• Add automated unit testing (started)
• Automatic Dask cluster configuration (started)
• Add ability for user to change file naming schemes and sort directories (this may be documentation on how to use existing API)
• (in reference to above) add case-specific uses of GenTS to user guide
• Integrate CLI features back into the package
• Create testing suite for verifying output
• Add contributor guide

GenTS Release v0.5.0

Features and Improvements ✨

• Consolidated all history file manipulations into HFCollection class
• Improved embarrassingly parallel workload generation
• Added documentation and ReadTheDocs page
• Added per-group time slicing
• Created framework for per-variable configuration
• Added preliminary integrity checks for time series files

Bug Fixes 🐛

• Fixed issue when sending large metadata task-graph to Dask cluster
• Updated timestamp strings to match old post-processing format
• Fixed issue with timestep detection resulting in seconds-level strings in time series file names
• Fixed error when handling multiple history files with different variable sets
• Removed redundant function calls
• Fixed non-thread safe parallel directory creation

Planned 📋

• Add default configurations for CESM1/2/3
• Add automated unit testing
• Report benchmarking results
• Integrate CLI features back into the package
• Create testing suite for verifying output
• Integrate with CUPiD

GenTS Release v0.4.1-beta

Bug Fixes 🐛

• Fixed bug in handling missing time bounds
• Improved metadata read performance by applying exclude path config

GenTS Release v0.4.0-beta

Features and Improvements ✨

• Implemented new modular design for more intuitive API
• Added filters and preliminary configuration system for saving and repeating post-processing tasks
• Added preliminary logging feature
• Created new class structure for handling meta data before generating time series
• Created preliminary system for checking file integrity

Bug Fixes 🐛

• Fixed bug in time sorting for concatenating history files
• Fixed issue with detecting aggregation dimension (which should always be time)
• Fixed time_bounds handling issue
• Fixed string formatting error to support older Python versions
• Fixed bug in generating names for output time series files

GenTS Release v0.3.0

Features and Improvements ✨

• Added time_bnds handling methods as alternative to using raw time values
• Moved attribute check further upstream to avoid needlessly loading history files for non-overwrite cases where timeseries already exist

Bug Fixes 🐛

• Fixed KeyError from grouping of existing timeseries with history files

GenTS Release 0.2.0

Features and Improvements ✨

• Added Command Line Interface (CLI)
• Implemented serial workflow if user doesn't want to use Dask
• Added verbosity level control and revamped internal logging mechanism
• Updated README to include new features and for easier reading
• Optimized metadata gathering parameters to decrease task data size
• Changed tag name to include 'v' prefix

Bug Fixes 🐛

• Fixed error with forcing serial mode during the build stage
• Fixed log messages not being stored in the class object (regardless of printing to the output stream)

GenTS Release 0.1.1

Features and Improvements ✨

• Package can be installed using PyPI (click here to view the PyPI page)!
• Added parameter for changing compression algorithm used by netcdf4 engine
• Integrated github workflow for automatic publishing to PyPI
• Added dynamic version attribute to keep track of what GenTS version was used to create the timeseries files
• Implemented dask bag when reading history file metadata to avoid overwhelming the scheduler with too many tasks
• Cleaned up documentation
• Simplified some arguments passing from build stage to run stage

Bug Fixes 🐛

• Fixed issue with year slicing that prevented last timeseries slice from being written

0.0.0 Initialization

Testing PyPI release workflow