complexheatmap-python 2.25.3

Python port of the R ComplexHeatmap package on a grid_py backend (tracks R ComplexHeatmap 2.25.3)

ComplexHeatmap-python

A Python port of the R ComplexHeatmap package (v2.25.3), built on grid_py — a pure-Python reimplementation of R's grid graphics system.

No matplotlib dependency. The entire rendering pipeline uses grid_py's Cairo backend, producing publication-quality output identical to R's grid device.

Features

• Heatmap with row/column clustering, splitting (k-means, factor), dendrograms, and custom color mapping
• HeatmapList for combining multiple heatmaps horizontally (+) or vertically (%), with synchronized row/column ordering and cross-heatmap body alignment
• 20 annotation functions: anno_barplot, anno_points, anno_lines, anno_mark, anno_text, anno_boxplot, anno_histogram, anno_density, anno_block, anno_image, and more
• HeatmapAnnotation with flexible sizing, gap control, and annotation name positioning
• Legends with continuous color bars and discrete icon legends, packable and positionable on all four sides
• oncoPrint for mutation landscape visualization
• UpSet plots for set intersection analysis
• Heatmap3D for 3D bar-style heatmaps with CIE LUV/HCL color space
• densityHeatmap and frequencyHeatmap for distribution visualization
• Global options via ht_opt() matching R's ht_opt$... interface
• Decoration API (decorate_heatmap_body, decorate_annotation, etc.) for post-draw customization

Installation

pip install complexheatmap-python

Or, for a local development checkout:

git clone https://github.com/Bio-Babel/ComplexHeatmap_python.git
cd ComplexHeatmap_python
pip install -e ".[dev]"

Quick Start

import numpy as np
from complexheatmap import Heatmap, color_ramp2

mat = np.random.randn(20, 10)
col = color_ramp2([-2, 0, 2], ["blue", "white", "red"])

ht = Heatmap(mat, name="example", col=col,
             row_km=3, column_title="My Heatmap")
ht.draw()

Combining Heatmaps

import grid_py as gp
from complexheatmap import Heatmap, color_ramp2, rowAnnotation, anno_barplot

np.random.seed(42)
mat1 = np.random.randn(20, 8)
mat2 = np.random.randn(20, 5)

ht1 = Heatmap(mat1, name="expr",
              col=color_ramp2([-2, 0, 2], ["green", "white", "red"]))
ht2 = Heatmap(mat2, name="cnv",
              col=color_ramp2([-1, 0, 1], ["blue", "white", "orange"]))

ht_list = ht1 + ht2 + rowAnnotation(
    bar=anno_barplot(np.random.randn(20), which="row"),
    width=gp.Unit(2, "cm"),
)
ht_list.draw(merge_legends=True)

Tutorials

Jupyter notebooks porting the official R ComplexHeatmap tutorials:

#	Notebook	Topic
1	Single Heatmap	Colors, clustering, splitting, annotations, labels
2	Heatmap Annotations	All 20 annotation types, sizing, decoration
3	A List of Heatmaps	Horizontal/vertical combination, row/column sync
4	Legends	Continuous/discrete legends, positioning
5	Heatmap Decoration	Post-draw customization via decorate API
6	OncoPrint	Mutation landscape visualization
7	UpSet Plot	Set intersection analysis
8	Integration	Combining with other analyses
9	Other High-Level Plots	densityHeatmap, frequencyHeatmap
10	3D Heatmap	Heatmap3D with bar projections
11	Genome-Level Heatmap	Chromosome-split genome visualization
12	Examples	Gene expression, measles, methylation
13	Other Tricks	Utility functions, text measurement, ht_opt

Testing

pip install -e ".[dev]"
pytest tests/