NEWS

---

nat.nblast 1.6.9 (2026-02-03)

• fix: stop using deprecated rgl::rgl.close()

nat.nblast 1.6.8 (2025-08-18)

• docs: add missing package anchors as requested by CRAN

nat.nblast 1.6.6

• Keep scale factor when normalising (#38)
• Doc fixes

nat.nblast 1.6.4

• pkgdown documentation
• use development version of nat package when doing install_github of nat.nblast
• spelling

nat.nblast 1.6.3

• fixes/examples for show_similarity
• allow score matrices to be passed as first param to nhclust (but issue a warning)
• change vignette engine to rmarkdown

nat.nblast 1.6.2 (2016-04-26)

• minor fixes for 3.3
• add vignette
• simplify travis setup

nat.nblast 1.6.1

• dev: fix test for different normalisation types

nat.nblast 1.6

• fix handling of non-square matrices in sub_score_mat
• minor doc fixes

nat.nblast 1.5 (2014-09-19)

• switch from importing nat to depending on it

nat.nblast 1.4

• use nabor package, not RANN, for finding nearest neighbours
• add functions for creating sparse score matrices, using spam package
• speed up extraction of diagonal elements for on-disk score matrices
• nblast now handles all combinations of neuron and dotprops objects

nat.nblast 1.3

• functions to create scoring matrices; see create_scoringmatrix for details.
• fix: really use nhclust's distfun argument
• new nblast_allbyall function
• add plyr support (gives parallelisation and progress bar options)

nat.nblast 1.2

• add nhclust, plot3d.hclust, subset.hclust to help cluster/plot based on nblast scores (all ported from flycircuit package).
• rename showSimilarity -> show_similarity
• dev: make sure tests run during check()
• test: update baselines given score matrix changes in 1.1

nat.nblast 1.1

• fix: make smat.fcwb the default scoring matrix (rather than smat_alpha) when nblast's UseAlpha parameter is FALSE (the default) and smat_alpha.fcwb the default when UseAlpha=TRUE.
• rename smat->smat.fcwb and smat_alpha->smat_alpha.fcwb to make it clear that these are defined in the FCWB template space (though they should work OK for other fly template brains that have an absolute physical scale)
• don't set options("nat.nblast.defaultsmat""), just query it.
• doc: basic package documentations