Suncal

Enhancements:

- Allow entry of Type A measurement data directly into Uncertainty interface
- Updated cost-based guardbanding to work with arbitrary distributions
- (Re)add Level of Confidence vs Coverage Factor dropdown in expanded uncertainty results
- Simplified data importing dialogs
- Moved some Risk calculation functions to Tools menu
- Added PFR Target guardbanding
- Show worst-case specific risk in risk calculation report
- Enabled dark mode switching

Bug Fixes:
- Fixed PFR calculation with single-sided limits
- Fixed guardbanding in Risk Curves tool
- Fixed potential issue calculating uncertainty with expression containing log of constant value

Other updates:
- Changed DataSets API to match other calculation types
- Removed deprecated UncertCalc class
- Upgraded UI to use QT6
- Updated for compatibility with Python 3.12

- Added inline help documentation
- Risk Analysis: Added calculation of conditional false accept risk and guardbanding based on conditional risk
- Risk Analysis: Added option to set process distribution using ITP
- Data Sets: Fixed errors when saving and displaying histograms with empty data
- Other minor bug fixes

API Changes - Risk Analysis (possibly breaking):

- Target PFA guardband calculation `risk.guardband` is now `risk.guardband.target`.
- `GB` parameter of `PFA_norm` and `PFR_norm` methods must be float. To calculate GB using a defined method, use methods in `risk.guardband_tur.*`.
- Removed `approx` keyword from risk `PFA`, `PFR` and `guardband.target methods`.
- `PFAR_MC` returns an additional element for Conditional PFA.

- Added re() and im() functions for ModelComplex calculations
- Fixed evaluation of constant model functions
- Corrected possible skew in histogram of Monte Carlo Distribution output plot
- Fixed issues with offset (eg temperature) unit conversion in uncertainty values
- Fixed issue when importing a distribution with measurement units
- Fixed entry of uncertainty level of confidence
- Replaced setup.py with setup.cfg

- Improve convergence of S2 Interval Calculation minimization problem

- Enabled coth and acoth functions
- Allow entering constant quantities with units into measurement model
- Restored sensitivity coefficients in GUM Derivation report
- Fixed GUI issues, including entering of chained uncertainty models, and reverse function names
- Fixed some reports when Monte Carlo results contained NaN or Inf values

UI changes:

- Implemented guided uncertainty wizard interface
- Added resizable splitters between sections in the window
- Updated ANOVA reports to better align with GUM examples

API changes (possibly breaking):

- Refactored API to use Model and ModelCallable classes, and deprecated UncertCalc class.
- Updated return values into Results objects and Reports objects for consistency across all calculation types.
- Updated for compatibility with Python 3.10 and 3.11. Minimum required version is 3.9.
- Makes some imports from subpackages
- Added API User Manual as Jupyter Notebooks in docs folder