About

fleur aims to provide some differences, but its key feature is the ability to create impressive and easy-to-reproduce plots with very few lines of code. It is well-suited for both exploratory data analysis and more goal-oriented analysis. More generally, fleur tries to:

• being super easy to use
  • automatically detect which test to use
  • make nice plots by default
  • minimalist API
• letting you control over everything: both from the statistics and dataviz point of view
• being more lightweight: it only relies on
  • matplotlib: for visualization
  • scipy: for statistics
  • narwhals: for data handling (fleur accepts all inputs that narhwals support: pandas, polars, pyarrow, cudf, modin).
• provide an extensive documentation with many examples.

Inspirations

fleur is highly inspired by the following projects:

• ggstatsplot: an R package that extends ggplot2 to add statistical details to plots.
• seaborn: the famous high-level interface of matplotlib for statistical data visualization.
• statannotations: a seaborn extension that adds statistical annotations.