pypubplot 0.1.0

Add your description here

pypubplot

pypubplot is an opinionated plotting library for single-cell analysis and academic publication. Built on top of ultraplot (a matplotlib wrapper), it provides publication-ready defaults: colorblind-friendly palettes, clean axis styling, and one-call export to PNG or SVG at ≥300 DPI.

Functionality

• Ridge plots — QC metric distributions grouped by cell type or sample.
• PCA scatter — Principal components with confidence ellipses and auto-labeled clusters.
• Embeddings (UMAP / t-SNE) — Categorical or continuous overlays with L-shaped axis stubs.
• Cell composition — Horizontal stacked-bar charts for subtype proportions.
• Violin plots — Grouped distributions with jitter dots and automatic legend handling.

All functions return (fig, ax) so you can fine-tune before saving.

Setup

Install from PyPI (once published) or directly from GitHub:

pip install pypubplot

Or with uv:

uv add pypubplot

Import the library:

import pubplot
from pubplot import plot_ridge, plot_pca, plot_violinplot
from pubplot import plot_embedding_categorical, plot_cell_composition

Usage

Ridge plot

Distributions of QC metrics (e.g. % mitochondrial) grouped by cell type or sample.

from pubplot import plot_ridge

plot_ridge(
    df,
    group_col="louvain",
    value_col="percent_mito",
    title="% mitochondrial — by cell type",
    xlabel="% mito",
    save_path="results/qc_pct_mt_by_cell_type",
)

PCA

Scatter plot of the first two principal components with confidence ellipses and auto-adjusted labels.

from pubplot import plot_pca

plot_pca(
    df,
    x_col="PC1",
    y_col="PC2",
    color_col="louvain",
    save_path="results/pca_pbmc3k",
)

UMAP / t-SNE

Categorical embeddings with cluster annotations and L-shaped axis stubs.

from pubplot import plot_embedding_categorical

plot_embedding_categorical(
    df,
    x_col="tSNE1",
    y_col="tSNE2",
    color_col="louvain",
    embedding_type="tSNE",
    save_path="results/tsne_pbmc3k",
)

from pubplot import plot_embedding_categorical

plot_embedding_categorical(
    df,
    x_col="UMAP1",
    y_col="UMAP2",
    color_col="louvain",
    embedding_type="UMAP",
    save_path="results/umap_pbmc3k",
)

Cell composition

Horizontal stacked-bar chart showing subtype proportions per sample.

from pubplot import plot_cell_composition

plot_cell_composition(
    composition_df,
    save_path="results/composition_8colors",
)

Violin plot

Grouped distributions with jitter dots. ≤ 8 groups get individual colours and a single-column legend; > 8 groups fall back to a uniform colour with x-axis labels.

from pubplot import plot_violinplot

plot_violinplot(
    df,
    group_col="louvain",
    value_col="NKG7",
    title="NKG7 — 5 cell subtypes",
    ylabel="Expression",
    save_path="results/violin_5subtypes",
)

Save format

By default all functions save to PNG at 300 DPI. Use save_fmt to control the output:

plot_pca(..., save_fmt="svg")   # SVG only
plot_pca(..., save_fmt="both")  # both PNG and SVG
plot_pca(..., save_fmt="png")   # PNG only (default)

Palettes

Built-in colourblind-friendly palettes are available directly:

from pubplot import PUBLICATION_PALETTE, OKABE_ITO
from pubplot import build_color_map, get_cell_composition_palette

Palette	Size	Use case
PUBLICATION_PALETTE	45	General-purpose, up to 45 groups
OKABE_ITO	8	Colorblind-safe categorical
CELL_COMPOSITION_PALETTE_8	8	Stacked bars (≤ 8 subtypes)
CELL_COMPOSITION_PALETTE_13	13	Stacked bars (≤ 13 subtypes)

For developers

Clone or download the repository on your machine. Make sure you have uv installed, then run:

uv sync

This will create the virtual environment and install all dependencies (including dev dependencies like ruff, mypy, and pre-commit).

To activate the pre-commit hooks:

uv run pre-commit install

All contributions are more than welcome. Feel free to open an issue or make a PR.

Author

Faouzi Braza