v0.7.4

Rendered Jupyter notebook tutorials to display output in docs. Resolves part of #84

v0.7.3

[0.7.3] Small patch

Improved

• Allowed the user to request files to be resampled upon loading. This is necessary for Mosqito metrics, which requires (and will itself resample) the audio files to be 48 kHz. The user can specify the desired sample rate in Binaural.from_wav() and higher level functions like AudioAnalysis.analyse_file, AudioAnalysis.analyse_folder.

Full Changelog: v0.7.1...v0.7.3

v0.7.1

Minor bug fixes

Full Changelog: v0.7.0...v0.7.1

v0.7.0

[0.7.0]

Complete refactoring of Soundscapy, splitting it into multiple modules (surveys, databases, audio, plotting), and improving the overall structure and functionality of the package. Also added more comprehensive documentation and test coverage.

General Changes

Added

• New soundscapy/surveys/survey_utils.py for shared utilities
  • Implemented PAQ enum for Perceptual Attribute Questions
  • Added return_paqs function for filtering PAQ columns
  • Created rename_paqs function for standardizing PAQ column names
• Centralized logging configuration in soundscapy/logging.py
  • Added support for environment variables to configure logging:
    • SOUNDSCAPY_LOG_LEVEL for setting log level
    • SOUNDSCAPY_LOG_FILE for specifying a log file
  • Implemented checks for Jupyter notebook environment in logging configuration
  • Added set_log_level function to allow dynamic adjustment of log level at runtime
  • Introduced global variable GLOBAL_LOG_LEVEL to manage log level across different environments
    • Implemented setup_logger function for initializing the logger with custom options
    • Set default logger to WARNING level with console output
    • Created get_logger function to retrieve the configured logger
• New processing module soundscapy/surveys/processing.py with enhanced functionality
  • Implemented ISOCoordinates and SSMMetrics dataclasses
  • Added calculate_iso_coords function for ISO coordinate calculations
  • Created add_iso_coords function to add ISO coordinates to DataFrames
  • Implemented likert_data_quality function for data quality checks
  • Added simulation function for generating random PAQ responses
  • Created ssm_metrics function for Structural Summary Method calculations
• Comprehensive docstrings and doctest examples in isd.py and satp.py
• New test cases in test_isd.py to cover refactored functionality

Changed

• Modified default logging level to WARNING for better control over log output
• Refactored isd.py to use new processing and survey utility functions
  • Updated load, load_zenodo, and validate functions
  • Refactored selection functions (select_record_ids, select_group_ids, etc.)
  • Updated describe_location and soundscapy_describe functions
• Refactored satp.py to align with new package structure
  • Updated load_zenodo and load_participants functions
  • Added doctest examples for all functions
• Modified __init__.py to initialize the logger when the package is imported
• Updated import statements across modules to use the new package structure
• Standardized function signatures and return types across all modules
• Changed to Rye as the dependency and environment manager for the project

Improved

• Enhanced error handling and input validation in database modules
  • Added type hints to all functions for better code readability and IDE support
  • Implemented more specific exception handling
• Optimized data processing functions for better performance
• Improved code organization and modularity
  • Separated concerns between data loading, processing, and analysis
• Enhanced documentation with more detailed explanations and examples
• Standardized coding style across all modules (using Black formatter)

Deprecated

• Removed remove_lockdown function in isd.py (redundant since the release of ISD v1.0)

Removed

• Eliminated redundant code and unused functions across modules

Fixed

• Resolved issues with inconsistent PAQ naming conventions
• Fixed bugs in ISO coordinate calculations and SSM metric computations
• Resolved issue where Jupyter notebooks were overriding the default log level

Security

• Implemented input validation to prevent potential security vulnerabilities

Development

• Implemented a more robust logging system using loguru
  • Added ability to easily change log levels for debugging and development
  • Enabled file logging for persistent log storage
• Enhanced test coverage for core functionality
• Added doctest examples to ensure documentation accuracy and serve as functional tests
• Implemented consistent error messages and logging across the package

Documentation

• Added comprehensive docstrings to all functions and classes
• Included usage examples in function docstrings
• Updated README with new package structure and usage instructions
• Created this CHANGELOG to track all significant changes to the project

Changes to Plotting Module

Code Structure:

• Split the original circumplex.py into multiple files: backends.py, circumplex_plot.py, plot_functions.py, stylers.py, and plotting_utils.py (implied).
• Introduced abstract base class PlotBackend and concrete implementations SeabornBackend and PlotlyBackend.

New Features:

• Added support for Plotly backend alongside Seaborn.
• Introduced CircumplexPlot class for creating and managing plots.
• Added StyleOptions dataclass for better style management.
• Implemented simple_density plot type.

Improved Customization:

• Created CircumplexPlotParams dataclass for better parameter management.
• Added more customization options for plots (e.g., incl_outline, fill, alpha).

Enhancements:

• Improved type hinting throughout the codebase.
• Added docstrings to classes and functions.
• Implemented PlotType and Backend enums for better type safety.

Refactoring:

• Moved plotting logic from functions to methods in backend classes.
• Simplified scatter and density functions by leveraging CircumplexPlot class.

Removed Features:

• Removed jointplot function (marked as TODO in CircumplexPlot class).

Constants and Utilities:

• Moved constants (e.g., DEFAULT_XLIM, DEFAULT_YLIM) to a separate utilities file.
• Created ExtraParams TypedDict for additional plotting parameters.

v0.6.0

Added

• Changed the name of the calculate_paq_coords to calculate_iso_coords to better reflect the function's purpose.
• Updated the formula for calculate_iso_coords to the more generalised form given in Aletta et. al. (2024).

v0.5.2

Minor bug fix. It seemed setuptools wasn't properly loaded as a scikit-maad dependency which would throw an error on import. Added setuptools to the dependencies.

Full Changelog: v0.5.0-alpha...v0.5.2

v0.5.0

What's Changed

• Refactored to remove pandas accessor methods by @MitchellAcoustics in #71

After using and teaching soundscapy more, I realised the pandas accessor approach actually just created more complexity and confusion. I liked the idea initially having the functionality as methods attached to the survey dataframe, but in the end it meant I needed to maintain twice the amount of code. It also ended up just confusing people.

So, I've deprecated the sspy and isd accessors. Now all of the functionality is included as standard functions. Some specific functions have been moved to the isd module if they really only make sense with data specifically in ISD format. This includes things like location_describe.

This will break your existing code but ideally it should be quite easy to fix and should make things much simpler moving forward.

The development package manager has also changed from poetry to pdm, which I find much nicer to work with. This has changed the pyproject.toml file. The dependencies have generally changed and been reduced. Soundscapy should now work with all python version > 3.9 and has also been updated to use pandas v2 which was recently released.

In addition, this fixes #69 and #65 which had a few errors and annoying warnings that were popping up.

Full Changelog: v0.4.4-alpha...v0.5.0-alpha

v0.4.4

What's Changed

• Update minimum python version and bug fixes by @MitchellAcoustics in #68
• Create python-package.yml by @MitchellAcoustics in #70

Full Changelog: v0.4.1-alpha...v0.4.4-alpha

v0.4.1

Minor release

• Added a progress bar for analysis functions
• Added a method for saving AnalysisSettings to a yaml file
• #Added additional tests for _AnalysisSettings

v0.4.0

What's Changed

• Switched the publication and dependency management to poetry
• Performed a major refactor to hopefully simplify use.
• Plotting has been separated out into 'circumplex' plots and 'likert' plots
• General purpose functions have been moved to utils
• Began building out separated support for different database.
• The dataframe accessor has been changed to 'sspy' to make it less specific to the ISD.
• Added docs using Sphinx by @MitchellAcoustics in #61

Full Changelog: v0.3.6-alpha...v0.4.0-alpha