- Fixed minor bugs related to parameter handling and data type coercion across various hedonic index calculation methods.
- Added an explicit error catch to the plotting function to prevent execution when
resting_points = TRUEis incorrectly passed.
- Major Architectural Refactoring: Extracted the linear model fitting and prediction logic from individual hedonic methods into centralized, high-performance internal helpers (
fit_hedonic_modelandformat_index_output) - Standardized internal data preparation and final index output formatting across Laspeyres, Paasche, Fisher, Time Dummy, and Repricing methods to guarantee strict consistency in output structures.
- Refactoring maintains exact API compatibility with
v1.1.0while significantly improving code maintainability and speed