NEWS
CanvasXpressBio 0.99.6
- Seed the demonstration-data helpers with
withr::with_seed() instead of
set.seed() so there is no set.seed() call in package code (clears the
BiocCheck warning), and tidy the related notes.
CanvasXpressBio 0.99.5
- New
makeDemoSE() builds the vignette's demonstration SummarizedExperiment
in one call, carrying the derived differential-expression table and the
per-cohort paired t-test p-values in its metadata(), so the vignette no
longer constructs data inline.
- New
makeBeeswarmDemo() returns a dense single-feature object for showing a
sina / beeswarm overlay at scale.
cxboxplot() gains points ("all", "sina", "jitter", "none") to pack
the observations into a beeswarm that stays legible with many points.
cxsurvival() gains tooltip to carry extra colData columns, so hovering a
step or a censored observation shows that subject's metadata.
- Vignette: build the dataset with
makeDemoSE(); document the enriched cell
hover and the view-reset control on the heatmap; surface per-subject metadata,
the censored-observation rug and the risk table on the Kaplan-Meier curve; and
add a beeswarm-at-scale example.
CanvasXpressBio 0.99.4
- Fix
cxvolcano(): the result matrix was transposed, so every gene collapsed
onto two plotted points. Genes are now the variables and the fold-change /
p-value are the axis columns, giving one point per gene.
cxvolcano() now colours points by regulation (up / down / not significant),
draws the fold-change and p-value threshold lines, and gains label,
label_genes, label_fc and label_p to label significant or named genes.
cxplot() gains overlays, var_overlays and cluster to draw colData /
rowData annotations as overlay tracks and to cluster rows and columns.
- New
cxsurvival() draws a Kaplan-Meier curve from colData.
- New
cxboxplot() draws a paired boxplot of one feature with per-subject
connecting lines, faceting and colouring from colData.
- Rewrote the vignette around a single
SummarizedExperiment shown as a
heatmap, volcano, Kaplan-Meier curve and paired boxplot, and added a section
motivating the package relative to existing Bioconductor visualization tools.
CanvasXpressBio 0.99.3
- Version bump to re-run the Bioconductor staging build (no code changes; 0.99.2 built
and passed checks but failed to publish).
CanvasXpressBio 0.99.2
- Vignette charts now render in the built HTML: require
canvasXpress (>= 1.70.3), which
ships a CanvasXpress library that BiocStyle's self-contained HTML post-processing no longer
corrupts (earlier versions were mangled on inlining, leaving the widgets blank). Verified by
rendering the vignette against the CRAN 1.70.3 library.
CanvasXpressBio 0.99.1
- Add maintainer ORCID iD to
Authors@R (BiocCheck note).
CanvasXpressBio 0.99.0
- Initial submission to Bioconductor.
cxplot() generic renders a SummarizedExperiment (and extending classes such as
SingleCellExperiment and DESeqDataSet) as an interactive CanvasXpress heatmap, with
colData as sample annotations and rowData as feature annotations. n= keeps only the
most variable features.
cxvolcano() draws a volcano plot from a differential-expression result table
(DESeqResults, limma::topTable(), or a data frame).