Changelog

All notable changes to EPyR Tools are documented here.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.3.5] - 2026-02-22

Fixed

  • FIX: Voigt derivative fitting now uses analytical Faddeeva function identities instead of numerical np.gradient. Derivatives are computed via w'(z) = -2z·w(z) + 2i/sqrt(pi) and w''(z) = -2w(z) + 4z²·w(z) - 4iz/sqrt(pi), providing smooth and accurate profiles required for curve_fit convergence.

  • FIX: Pseudo-Voigt derivative fitting now correctly passes derivative and phase parameters to the underlying lshape() function. Previously these parameters were ignored and a false warning was emitted.

  • FIX: lshape() Gaussian/Lorentzian components now delegate to the canonical gaussian() and lorentzian() implementations, replacing a buggy internal Dawson function approximation that returned complex values when phase was nonzero.

  • FIX: fit_epr_signal() initial parameter estimation for derivative signals. New heuristics for 1st and 2nd derivative data correctly estimate center (zero-crossing), width (peak separation), and amplitude (model-calibrated scaling).

  • FIX: plot_2d_map() no longer crashes with recent matplotlib (3.8+) due to tight_layout / colorbar incompatibility. Uses constrained layout instead.

  • FIX: baseline_polynomial_2d() meshgrid shape mismatch for 2D data. np.meshgrid(x[0], x[1]) now produces arrays matching y.shape.

Added

  • NEW: vmin and vmax parameters for plot_2d_map() to control the color scale range.

Changed

  • DEFAULT: Default line width set to 0.75 for plot_1d() and plot_2d_waterfall() for cleaner plots.

  • DEFAULT: plot_2d_waterfall() default max_traces reduced from 50 to 20.

  • DEFAULT: JPG export (save_to_jpg) now uses 200 DPI (was 150), RdBu_r colormap for 2D maps, and symmetric color scale based on 98th percentile of absolute values.

Compatibility

  • Backward compatible: all existing function signatures preserved.

  • New parameters use default values matching previous behavior where possible.

[0.3.1] - 2026-02-03

Added

  • NEW: Automatic file extension detection for EPR data loading

  • NEW: fit_baseline option in fit_epr_signal() for affine baseline correction

Fixed

  • Fixed filename parsing for files with multiple dots in name

[0.3.0] - 2025-11-02

Changed

  • MAJOR: Migrated 398 print statements to centralized logging system (95.7% coverage)

    • Refactored 18 core modules with structured logging infrastructure

    • logger.info() for user-facing messages and progress updates

    • logger.warning() for errors, issues, and missing dependencies

    • logger.debug() for technical details and internal operations

    • Modules updated: eprload, frequency_analysis, conversions, cli, units, baseline (multiple), signalprocessing/apowin, fair (multiple), physics/constants, lineshapes (fitting, convspec), eprplot, isotope_gui

  • IMPROVEMENT: Enhanced code quality with professional logging standards

    • Centralized logging configuration via logging_config module

    • Consistent log format with timestamps, module names, and severity levels

    • Better traceability for debugging and production monitoring

    • Proper log level hierarchy (DEBUG < INFO < WARNING < ERROR)

Infrastructure

  • All core modules now use get_logger(__name__) pattern for consistent logging

  • Logging properly configured for both CLI commands and library usage

  • No breaking changes - all functionality preserved and tested

Compatibility

  • Backward Compatible: All existing functionality preserved

  • Test suite confirms no regressions (100+ tests passing)

  • API unchanged - only internal output mechanism modified

  • Users can configure logging levels via standard Python logging configuration

[0.2.5] - 2025-10-07

Added

  • NEW: return_type parameter in eprload() for selective component extraction

    • return_type="default": Return signal as-is (backward compatible)

    • return_type="real": Extract only real part

    • return_type="imag": Extract only imaginary part

  • NEW: Advanced 2D FFT analysis module (analyze_frequencies_2d)

    • Row-by-row 1D FFT mode for 2D Rabi oscillations

    • Full 2D FFT mode for HYSCORE measurements

    • Automatic time unit detection (ns, μs, ms, s)

    • DC offset removal and apodization windowing

    • Centered frequency spectrum (fftshift for positive and negative frequencies)

    • Four-panel visualization: original, FFT linear/log, phase spectrum

Changed

  • DEFAULT CHANGE: eprload() parameter plot_if_possible now defaults to False

    • Previous default was True (automatic plotting)

    • New default is False for cleaner programmatic use

    • Explicitly set plot_if_possible=True if automatic plotting is desired

  • REFACTORED: Signal processing module (frequency_analysis.py)

    • Extracted helper functions to eliminate code duplication

    • _detect_time_units(): Unified time unit detection

    • _convert_to_display_freq(): Consistent frequency conversion

    • _remove_dc_offset(): Standardized DC removal

    • _apply_window(): Centralized apodization

    • Reduced code duplication by ~150 lines

    • Single source of truth for common operations

Documentation

  • Updated README with v0.2.5 features and contributor information

  • Added institutional logos (LogoL.png, LogoR.png)

  • Created comprehensive RELEASE_NOTES_v0.2.5.md

  • Updated version badges to 0.2.5 across all documentation

  • Enhanced API documentation for new parameters

Testing

  • Added tests for return_type parameter in eprload()

  • Validated 2D FFT analysis with synthetic EPR data

  • Verified backward compatibility for all existing workflows

  • Confirmed refactored code produces identical results

[0.2.0] - 2025-09-27

Added

  • NEW: Stable core functionality with validated test suite

  • NEW: FAIR data conversion with 235+ parameter mappings

  • NEW: Plugin system for extensible file formats

  • NEW: Configuration management with environment variable support

  • NEW: Memory optimization and performance tracking

Changed

  • IMPROVEMENT: Updated to stable release status

  • CLEANUP: Removed 29MB of unnecessary files

  • UPDATE: Comprehensive documentation overhaul

Fixed

  • Core test suite passing at 100%

  • Memory management improvements

  • Dependency verification

[0.1.6] - 2025-09-12

Added

  • NEW: Comprehensive testing protocol with 4 levels (SMOKE, STANDARD, DEEP, SCIENTIFIC)

  • NEW: Deep testing infrastructure with automated test runners

  • NEW: Complete lineshape analysis system with mathematical validation

  • NEW: Performance benchmarking for all core functions

  • NEW: Scientific accuracy validation against NIST standards

Changed

  • BREAKING: Removed all numbered notebooks from examples/ for cleaner project structure

  • IMPROVEMENT: Removed all emojis from documentation for professional appearance

  • UPDATE: Version updated to 0.1.6 across all configuration files

  • ENHANCEMENT: Updated documentation with comprehensive testing information

  • FIX: Fixed pseudo_voigt parameter handling in lineshape_class

Removed

  • Removed numbered example notebooks (01_, 04_, 05_, 06_, 07_, 08_, 09_, 10_)

  • Cleaned emojis from all markdown documentation files

Fixed

  • Fixed voigtian function calls with proper parameter structure

  • Corrected lineshape function parameter handling

  • Updated all version references to 0.1.6

[0.1.2] - 2025-09-05

Removed

  • BREAKING: Removed epyr.sub.baseline2.py - deprecated duplicate baseline functions

  • BREAKING: Removed epyr.sub.processing2.py - deprecated duplicate processing functions

  • Cleaned up duplicate code and imports

Changed

  • Updated package imports to remove references to deleted modules

  • All baseline correction functions now available through epyr.baseline module

  • Streamlined package structure for better maintainability

Fixed

  • Fixed import issues in Getting Started notebook

  • Consolidated all data files into single examples/data/ directory

  • Fixed complex data handling in notebooks

  • Updated path detection for cross-platform compatibility

Documentation

  • Updated README with version badge

  • Created comprehensive Getting Started notebook with real data examples

  • Added proper error handling and troubleshooting in notebook

  • Updated all version references

Migration Guide

If you were using the removed modules:

# OLD (no longer works)
from epyr.sub.baseline2 import baseline_polynomial
from epyr.sub.processing2 import baseline_polynomial

# NEW (use instead)
from epyr.baseline import baseline_polynomial

[0.1.1] - 2025-09-04

Added

  • Comprehensive README with professional documentation

  • Setup.py for pip package installation

  • Example notebooks and tutorials

  • FAIR data conversion capabilities

  • Advanced plotting functionality

Fixed

  • Various import and compatibility issues

  • Documentation generation

  • Test coverage improvements

[0.1.0] - 2025-09-01

Added

  • Initial release

  • EPR data loading (BES3T and ESP formats)

  • Basic baseline correction

  • Constants and physical parameters

  • Isotope GUI application

  • Basic plotting capabilities

Features by Version

Core Data Loading

Version 0.1.0+

  • Load BES3T (.dsc/.dta) files from modern Bruker spectrometers

  • Load ESP (.par/.spc) files from legacy Bruker systems

  • Automatic format detection and parameter extraction

  • Support for both 1D spectra and 2D datasets

  • Complex data handling for pulsed EPR experiments

Version 0.1.1+

  • Enhanced error handling and validation

  • Cross-platform file path compatibility

  • Improved parameter parsing and units

Version 0.1.2+

  • Consolidated data directory structure

  • Enhanced 2D data support in examples

  • Better complex data visualization

FAIR Data Conversion

Version 0.1.1+

  • Convert to CSV format with metadata headers

  • Export to JSON with complete parameter documentation

  • Save as HDF5 for efficient large dataset storage

  • Preserve all experimental metadata

  • Cross-platform compatibility

Advanced Baseline Correction

Version 0.1.0+

  • Polynomial baseline correction (orders 0-5)

  • Signal region exclusion from baseline fitting

  • 1D spectrum processing with validation

Version 0.1.1+

  • Exponential decay models (single and stretched)

  • Improved parameter estimation algorithms

  • Better error handling for edge cases

Version 0.1.2+

  • Streamlined API through unified epyr.baseline module

  • Removed duplicate and deprecated functions

  • Enhanced documentation and examples

Visualization Tools

Version 0.1.0+

  • Basic EPR spectrum plotting

  • Parameter display and annotation

  • Export to common image formats

Version 0.1.1+

  • 2D spectral maps with customizable color schemes

  • Interactive plotting capabilities

  • Publication-quality output options

Version 0.1.2+

  • Enhanced 2D data visualization in examples

  • Complex data magnitude plotting

  • Improved axis labeling and units

Documentation and Examples

Version 0.1.0+

  • Basic API documentation with Sphinx

  • Simple usage examples

  • Core function docstrings

Version 0.1.1+

  • Comprehensive README with installation guide

  • Professional Sphinx documentation theme

  • Tutorial notebooks for interactive learning

Version 0.1.2+

  • Complete ReadTheDocs.io integration

  • Detailed installation and contribution guides

  • Enhanced example scripts and notebooks

  • Cross-referenced API documentation

Development Infrastructure

Version 0.1.0+

  • Basic package structure

  • Essential dependencies (NumPy, Matplotlib, SciPy)

  • Git version control

Version 0.1.1+

  • Standard setup.py for pip installation

  • Development dependencies and tools

  • Code quality checks (Black, isort, flake8)

Version 0.1.2+

  • Pre-commit hooks for automated code quality

  • Comprehensive test suite (44 tests)

  • Modern packaging with pyproject.toml

  • ReadTheDocs configuration for automatic documentation builds

Deprecation Notices

Version 0.1.2

The following modules were removed as they contained duplicate functionality:

  • epyr.sub.baseline2 - Use epyr.baseline instead

  • epyr.sub.processing2 - Functionality merged into main modules

These modules were deprecated since version 0.1.1 and have been removed to streamline the package structure.

Future versions will maintain backward compatibility for the public API in epyr.baseline, epyr.eprload, and epyr.fair modules.

Upgrade Path

From 0.1.1 to 0.1.2

  1. Update imports:

    # Change this:
from epyr.sub.baseline2 import baseline_polynomial

# To this:
from epyr.baseline import baseline_polynomial

  2. Update data paths: If you were using separate BES3T/ and ESP/ directories, consolidate data files into a single directory.

  3. Check examples: Updated example scripts now handle both 1D and 2D data automatically.

From 0.1.0 to 0.1.1

No breaking changes - all existing code should continue to work.

New features can be adopted incrementally:

  • Use epyr.fair module for data conversion

  • Try new example notebooks in examples/notebooks/

  • Install with pip install -e . for easier development

Breaking Changes by Version

Version 0.1.2

  • Removed epyr.sub.baseline2 module

  • Removed epyr.sub.processing2 module

  • Changed example data directory structure

Version 0.1.1

No breaking changes from 0.1.0.

Version 0.1.0

Initial release - no previous versions to break compatibility with.