
Articles
Get started
Guides
Deeper looks at what HeatmapR can do, with every figure drawn as the site is built.
- Building heatmaps in the browser
heat_map_app() builds heatmaps, dot plots and correlation heatmaps from a file or from data in R without writing code, and writes the R code that draws them again.
- Ordering rows and columns
Dendrogram, optimal leaf and spectral orders, how they work and when to use each, measured on clusters, a trajectory and a cycle.
- Large heatmaps
Hundreds of thousands of rows: raster drawing, two-stage clustering, k-means splits, automatic cuts, labels and zoom panels, threads and memory.
- Single-cell data
Heatmaps, dot plots and markers straight from SingleCellExperiment, Seurat, flowFrame and sparse matrices: features by name, groups and annotations from metadata, downsampling, sparse summaries, marker ranking, cytometry and trajectories.
- More than colour
Uncertainty, two values in a cell, distributions inside cells and beside the heatmap, metadata annotations, honest colour scales and correlations.
- Working with heatmaps
Redrawing and saving heatmaps, extracting their orders and clusters, tidy data for ggplot2, heatmaps beside ggplot2 plots and alternative text for screen readers.
Performance
- Speed comparison
The time and memory HeatmapR, ComplexHeatmap, pheatmap, gplots, stats and ggplot2 need to draw and cluster heatmaps from 10,000 to 9 million cells, and of clustering and leaf ordering on their own.