- 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.
- 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).
- 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()).