REPS 1.1.1

Bug Fixes

• 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 = TRUE is incorrectly passed.

Internal Improvements

• Major Architectural Refactoring: Extracted the linear model fitting and prediction logic from individual hedonic methods into centralized, high-performance internal helpers (fit_hedonic_model and format_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 v1.1.0 while significantly improving code maintainability and speed

REPS 1.1.0

REPS 1.1.0

Major Changes

• Renamed calculate_price_index to calculate_hedonic_index.

Performance Improvements

• Accelerated the HMTS method for faster execution.

Bug Fixes

• Applied bug fixes for the Rolling Time Dummy and Time Dummy methods.

REPS v1.0.0

Note: This release marks the rebranding and official first major release of the package.
The package was previously known as cbsREPS (version 0.1.0) and is now renamed to REPS.

---

Major Features

New Index Methods Supported in calculate_price_index()

• Added support for three additional index calculation methods:
  • Repricing method (hedonic geometric adjustment)
  • Time Dummy index
  • Rolling Time Dummy index
• These methods can now be called individually or together via the method argument.
• Example:

  calculate_price_index(..., method = c("fisher", "laspeyres", "repricing"))

Multi-Method Execution and Output

• When passing multiple methods, the function returns a named list of outputs.
• Enables seamless comparison of different price index methodologies in a single step.

New Plotting Utilities

• plot_price_index() introduced for quick visualization of index results.
  • Automatically detects multi-method input.
  • Clearly labeled plots with period breaks for time-based readability.

---

Regression Diagnostics

New Function: calculate_regression_diagnostics()

• Computes diagnostic statistics for log-linear hedonic models per period:
  • Adjusted R²
  • Shapiro-Wilk normality test
  • Durbin-Watson test for autocorrelation
  • Breusch-Pagan test for heteroscedasticity

New Function: plot_regression_diagnostics()

• Visualizes diagnostics over time in a 3×2 grid.
• Includes threshold markers to easily spot anomalies or violations of assumptions.

---

Bug Fixes and Compatibility

• Improved input validation in validate_input():
  • Clear distinction between errors and warnings.
  • Accepts period formats like "202101" and "2020Q1".
• Removed dependency on assertthat; replaced with native R error handling.
• Guardrails added to prevent invalid combinations (e.g., resting_points = TRUE with multiple methods).

---

Documentation and Testing

• Comprehensive updates to documentation, examples, and function manuals.
• New vignette: calculate_regression_diagnostics showcasing diagnostics workflow.
• Expanded test coverage across all new features to ensure stability and correctness.

---

REPS 1.0.0 sets the stage for scalable, reproducible price index analysis. Whether you’re working with Fisher, Laspeyres, or regression-based methods, this version provides you with the tools needed to compute, compare, and interpret indices with ease and statistical rigor.

📦 cbsREPS 0.1.0 — Initial Release

calculate_price_index: Introduces a unified interface for computing price indices, using following methods:

• Laspeyres index
• Paasche index
• Fisher index
• HMTS (Hedonic Multilateral Time series re-estimation with splicing)

Full Changelog: v0.1.0...v0.1.0