v0.1.0

First public release. Seventeen visualization components, each shipping from one
source bundle to JavaScript, R and Python.

Added

• Shared core (plotomics/core). A single TypeScript core defining the
  component contract, theming, colour scales, binary column transport and export
  helpers, consumed by every visualization so behaviour stays consistent across
  languages. Includes the viridis and rdbu ramps plus the ltc sequential
  and ltcdiv diverging ramps. Ships as a subpath of the one npm package rather
  than a second package, so there is no version skew between core and components.
• Tri-language wrapper architecture. Each component ships from one source
  bundle to three targets: an npm package (plotomics), a Python
  anywidget widget (plotomics on PyPI, for Jupyter,
  JupyterLab, marimo, Colab, VS Code, Shiny for Python and Streamlit) and an R
  htmlwidget (plotomics, for the RStudio Viewer,
  R Markdown, Quarto and Shiny). Glob-discovered build entries let new components
  land without editing any central registry.
• Binary column transport. Numeric data reaches the browser as a binary
  buffer rather than JSON, which is what keeps hundreds of thousands of points
  interactive.

Expression and abundance

• Volcano plot. GPU-accelerated differential-expression scatter with
  significance thresholds.
• Expression heatmap. Large-scale WebGL expression matrix viewer.
• Clustered heatmap (clustermap). Heatmap with row and column dendrograms and
  hierarchical clustering.
• Marker gene dot plot. Features by groups, with dot area proportional to the
  fraction of the group expressing and colour carrying the level.
• Stacked violin. One row per feature, one violin per group, so a marker
  panel reads down the page. violin_density() computes the densities in R.

Single-cell and spatial

• Embedding (UMAP/t-SNE) viewer. GPU-accelerated scatter for single-cell and
  dimensionality-reduction embeddings. aspect = "equal" gives both axes the
  same units per pixel, for PCA scores and anything else whose axes share units;
  point_scale_mode chooses how points shrink as you zoom out; padding keeps
  the outermost points off the canvas border. A factor color column in R, or a
  pandas Categorical in Python, pins the legend order and colour assignment and
  keeps unused levels, the way drop = FALSE does in ggplot2.
• Spatial tissue map. Measurements at their real slide coordinates over the
  histology image, with the image and the spots sharing a single fit so they
  cannot drift apart on resize or on a high-DPI display.

Cohort and variant genomics

• Oncoplot (OncoPrint). The cohort alteration landscape: a gene by sample
  grid of categorical alteration classes, with a per-sample mutation-burden
  barplot, a per-gene frequency barplot and optional clinical annotation strips.
  oncoplot_memo_sort() produces the conventional column order.
• Protein domain lollipop. Variants along a protein over its domain
  architecture, with head area proportional to recurrence and an optional
  post-translational modification track.
• Kaplan-Meier curve. A right-continuous step function per stratum with
  censoring ticks, confidence bands and a number-at-risk table aligned to the
  same time grid. Accepts a survival::survfit object directly in R.
• Categorical profile. A grouped bar profile with coloured header blocks,
  built for the 96-context mutational signature layout and general enough for any
  ordered categorical profile that groups into runs.

Sets, hierarchies and networks

• UpSet plot. Set intersections as a bar chart over a membership matrix, for
  the many-set case where a Venn diagram is not drawable.
  upset_intersections() computes exclusive intersections, so the columns sum to
  the union rather than double-counting.
• Gene treemap. Hierarchical gene and category treemap.
• Network graph. Large-graph viewer built on
  Sigma v3 and
  graphology, with directed edges, per-edge
  colour and drop reporting. In Shiny, clicking a node pushes its id to
  input$<outputId>_selected.

Genome and chromatin

• Hi-C contact matrix. WebGL contact-map viewer with level-of-detail tiling,
  for chromatin interaction data.
• igv.js genome viewer. Embedded igv.js genome browser.
• Gosling genome viewer. Declarative genomics visualization via
  Gosling.js.

Documentation

• docs/motivation.md covers why the project exists, its
  non-goals, how it compares with ComplexHeatmap, maftools, survminer, UpSetR,
  Seurat and scanpy, and when to reach for one of those instead.
• docs/architecture.md describes the component contract
  and the transport layer.
• Reference sites are generated per language: pkgdown for R, pdoc for Python.