# depictr

A unified, colourblind-safe toolkit for publication-ready statistical
visualisation, built on [plotnine](https://plotnine.org). It is the Python
sibling of the [depictr R package](https://github.com/pablobernabeu/depictr).

![A grouped density of response times by condition, in the colourblind-safe Okabe-Ito palette](https://raw.githubusercontent.com/pablobernabeu/depictr-py/main/images/readme-distribution.png)

```{toctree}
:maxdepth: 1
:hidden:

Getting started <getting-started>
Gallery <auto_examples/index>
API reference <api>
About <about>
```

## What it gives you

Python already has an excellent plot for almost any statistical task, but they
live in different packages with different defaults, different APIs and different
colour schemes. depictr gives the whole workflow one theme, one colourblind-safe
palette and one calling convention, and returns plotnine objects you can keep
extending with `+`. Where a specialist package already computes a quantity well,
depictr delegates to it and redraws the result under the shared theme.

The default palette is the Okabe-Ito set, and that choice is checked rather than
asserted: the package ships a Machado-2009 colour-vision-deficiency simulator and
a CIE-Lab distance test that confirms the colours stay distinguishable under each
deficiency.

## Installation

depictr is on [PyPI](https://pypi.org/project/depictr/):

```bash
pip install depictr            # core (plotnine, pandas, numpy)
pip install depictr[all]       # plus the optional computation back-ends
```

The classification, model and survival plots delegate to scikit-learn,
statsmodels and lifelines; each is an optional dependency, so the core stays
light.

## A first plot

```python
import depictr as dp

ld = dp.lexical_decision()
dp.explore_distribution(ld, "RT", group="condition", kind="both")
```

The [gallery](auto_examples/index) works through every family with the plots
rendered, and the [API reference](api) documents each function.