Skip to content

REPS v1.0.0

Choose a tag to compare

@vivekag7 vivekag7 released this 30 Jul 08:45
· 45 commits to develop since this release

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.