LoginSign Up
← Back to all sparks
G

geocomplexity

INFRA · APIS
Velocity0.0

Mitigating Spatial Bias Through Geographical Complexity

github.com/ausgis/geocomplexity ↗

A spatial complexity package that shipped its method, then went quiet

spatial-statisticsr-packagegeographically-weighted-regressionrastermaintenance
◆Current state
geocomplexity computes geographical complexity from spatial dependence and configuration similarity across both vector and raster data, and uses it to build spatial weight matrices and a complexity-aware geographically weighted regression. That capability arrived complete in the 0.1.0 release of September 2024. The three releases since contain no functional change: a citation file, a dependency trim, one function moved out to a sibling package, and a maintainer surname correction.
◆Where it's heading
The package sits inside Wenbo Lyu's spatial statistics family, where shared functionality migrates into the common sdsfun package rather than being duplicated across dependents. moran_test left geocomplexity for sdsfun in 0.2.0, which is the same consolidation pattern visible across the author's other packages. What remains here is the method-specific surface, and it has not changed in eighteen months.
◆Prediction
The entries give no signal of planned functional work; on this pattern the next release is as likely to be metadata or another function migration to sdsfun as anything user-visible.

◆Recent moves

  1. 4mo ago

    Maintainer surname spelling corrected for pinyin compliance

    Changes the maintainer surname from Lv to Lyu for pinyin compliance. Metadata only, and the sole release in eighteen months - which is itself the most informative thing about the package's current state.

    View source ↗
  2. 1y ago

    Package citation added

    Adds a citation entry so the package can be cited in publications. Standard practice for an academic method package and nothing more.

    View source ↗
  3. 1y ago

    moran_test migrates out to the shared sdsfun package

    Moves moran_test into sdsfun and drops tidyverse from suggested dependencies. This is the author's consolidation pattern rather than a loss of capability: general spatial functions collect in sdsfun so the dependent packages share one implementation instead of each carrying a copy.

    View source ↗
  4. 1y ago

    Geographical complexity measures and complexity-weighted GWR

    The release that defines the package: complexity computation over vector and raster data, spatial weight matrices derived from that complexity, and a geographical complexity form of geographically weighted regression. Everything published since has been packaging around this unchanged core.

    View source ↗