cinderplot-examples

Example datasets and the regression test suite for cinderplot. Kept out of the code repository so that stays small and buildable; the data lives here.

The documentation site and gallery HTML live in the code repo under docs/ (published to GitHub Pages at zhou-lab.github.io/cinderplot/). Its build script reads the datasets here from a sibling checkout — see docs/build.py in the cinderplot repo.

The rendered gallery figures are served from this repo's Pages site to keep the code repo free of growing binary assets: docs/figs/*.svg (cinderplot) and docs/figs/*.png (ggplot2 reference), published at zhou-lab.github.io/cinderplot-examples/figs/… and linked by the gallery. Enable it under Settings → Pages → Deploy from a branch → main / docs.

Layout

dir	what
data/	small CSV datasets (mtcars, expr, …) used by the tests and quick demos
genome/	whole-genome CNV example (K562): bins, segments, hg38 seqinfo + cytoband
tracks/	locus track-browser example (synthetic placeholder): coverage, genes, peaks, loops
region/	sesame visualizeRegion target (real ADA/HM450 data): a genome-anchored beta heatmap + gene models — see its README for the co-dev spec
tests/	regression suite (test.sh) + cluster_check.c

Each subdirectory has its own README with the exact commands.

Running the examples

These assume cinderplot is on your PATH (built from the code repo and installed, or make install). Sample data lives in data/. Reference genome annotation (cytoband, seqinfo, gene models) is read straight from a genome repo: cinderplot decompresses gzip/bgzip and region-queries a tabix index in memory, so point those arguments at your local copy (e.g. ~/repo/genomes/hg38):

cinderplot data/k562.bins.tsv \
  '... + scale_x_genome("/path/to/genomes/hg38/seqinfo.tsv.gz")
       + ideogram("/path/to/genomes/hg38/cytoband.tsv.gz") ...' -o k562.pdf

Running the tests

tests/test.sh needs the built binary. By default it looks for a sibling checkout of the code repo at ../cinderplot/cinderplot; override with CINDERPLOT:

CINDERPLOT=/path/to/cinderplot sh tests/test.sh

tests/cluster_check.c is a standalone verifier that links against the cinderplot sources; compile it from within the code repo (it needs include/cinderplot.h and cluster.c/csv.c).