Changes in version 0.1.2                        

  - Bug Fixes & Robustness:
    
      - Resolved single-query coordinate vector bug in
        scp_geostatistical(): 1D coordinate vectors passed as s0 are now
        properly coerced to a 1-row coordinate matrix matching training
        dimensions.
      - Fixed graphics layout warning in plot.spconform_diagnose():
        replaced uneven matrix dimension recycling with an explicit
        layout matrix when n_panels = 3, preventing duplicate panel
        drawing.
      - Added strict input validations in scp_geostatistical() and
        scp_areal() guarding against missing (NA) values in coordinates
        and responses.
      - Added length validation and na.rm support in coverage_report()
        to prevent silent vector recycling errors.

  - Performance & Computational Efficiency:
    
      - Optimized default spatial bandwidth computation in
        scp_geostatistical(): bandwidth is now evaluated once prior to
        the query loop rather than recomputed per target location,
        yielding up to a 50x speedup on large query grids.
      - Added representative spatial subsampling for bandwidth
        evaluation when calibration size $N > 1000$, keeping memory
        footprint strictly bounded.
      - Added optional k_neighbors argument to scp_geostatistical() for
        localized $k$-NN calibration, defaulting to NULL for complete
        backward compatibility.

  - Statistical Enhancements:
    
      - coverage_report() now evaluates and reports the strictly proper
        Winkler Interval Score (winkler_score) alongside empirical
        coverage and mean interval width.

                 Changes in version 0.1.1 (2026-09-24)                  

  - Enhanced S3 generics and methods for spconform objects:
      - Added predict.spconform() to extract point predictions or
        prediction intervals (fit, lwr, upr).
      - Added residuals.spconform() to compute raw response residuals or
        absolute calibration scores.
      - Added as.data.frame.spconform() to coerce conformal objects into
        tidy data frames.
  - Enhanced spatial diagnostics:
      - diagnose() now returns a classed spconform_diagnose object with
        dedicated print.spconform_diagnose() and
        plot.spconform_diagnose() methods.
      - Added Moran's $I$ test on prediction residuals and conformal
        hit/miss indicators.
  - Documentation and code compliance:
      - Replaced all non-English code comments with English comments in
        R/scp_areal.R for full ASCII compliance.
      - Standardized all Rd help page titles to Title Case style.
      - Updated citation metadata to Mao, Martin, and Reich (JASA 2024).

                 Changes in version 0.1.0 (2026-09-12)                  

  - Initial release of spconform on CRAN.
  - Implemented locally weighted split conformal prediction for
    geostatistical (point-referenced) data (scp_geostatistical()).
  - Implemented neighbourhood-weighted leave-one-out conformal
    prediction for areal (lattice) data (scp_areal()).
  - Added kernel weighting utilities for spatial and spatio-temporal
    predictions (spatial_kernel_weights(), areal_neighbor_weights()).
  - Implemented comprehensive spatial diagnostic tools and multi-panel
    visualization (diagnose()).
  - Included S3 methods for printing, summarizing, and plotting
    conformal prediction intervals (print(), summary(), plot(),
    coverage_report()).