ArviZ-plots

1.0.0 · active · verified Thu Apr 16

ArviZ-plots is a new Python library (v1.0.0) designed to provide ready-to-use and composable plots specifically for Bayesian workflow. It offers a flexible and modular plotting interface, building upon ArviZ's `InferenceData` object for modern visualization needs. As a `1.0.0` release, it aims for API stability in future patch versions.

Common errors

```
ModuleNotFoundError: No module named 'arviz_plots'
```
cause The arviz-plots library is not installed in your current Python environment.

fix Run `pip install arviz-plots` to install the package.
```
TypeError: plot_pair() missing 1 required positional argument: 'data'
```
cause A plotting function like `plot_pair` was called without providing the necessary `InferenceData` object.

fix Pass an `arviz.InferenceData` object as the first argument, e.g., `plot_pair(idata)`.
```
ERROR: Package 'arviz-plots' requires a different Python version: 3.12 or later but 'X.Y' is specified.
```
cause Your current Python version does not meet the minimum requirement (Python 3.12).

fix Upgrade your Python environment to version 3.12 or newer. Consider using a virtual environment with the correct Python version.

Warnings

gotcha ArviZ-plots is designed to work exclusively with `arviz.InferenceData` objects. Providing other data structures will lead to errors.
Fix: Ensure your data is loaded or converted into an `arviz.InferenceData` object before passing it to ArviZ-plots functions. Use `arviz.InferenceData` or `arviz.from_pymc`/`from_stan` etc.
gotcha ArviZ-plots is not a direct, drop-in replacement for *all* `arviz.plot_*` functions. While it offers similar plots, its API is different and focuses on composability. Not all ArviZ plots are yet implemented in arviz-plots, or they may have different parameters.
Fix: Consult the `arviz-plots` documentation for available functions and their specific usage. Do not assume a 1:1 migration path from `arviz.plot`.
breaking ArviZ-plots requires Python 3.12 or newer. Installing or running the library on older Python versions will fail.
Fix: Upgrade your Python environment to version 3.12 or newer. Use `pyenv` or `conda` to manage multiple Python versions if necessary.
breaking As a new `1.0.0` release, while patch releases will aim for API stability, future *major* versions (e.g., `2.0.0`) are likely to introduce significant API changes to refine the composable plotting interface and integrate new features.
Fix: Pin your `arviz-plots` dependency to `~1.0` if strict API stability is required for a production environment. Review migration guides for future major versions.

Install

pip install arviz-plots Install stable release

Imports

plot_pair
```
from arviz_plots import plot_pair
```
plot_trace
```
from arviz_plots import plot_trace
```
Plot
```
from arviz_plots.base import Plot
```
Represents a composable plot object, typically returned by plotting functions.

Quickstart

This example loads sample Bayesian inference data using ArviZ and then generates a pair plot using ArviZ-plots, demonstrating the core `plot_pair` function.

import arviz as az
from arviz_plots import plot_pair

# Load example InferenceData from ArviZ
idata = az.load_arviz_data("centered_eight")

# Create a basic pair plot
plot_pair(idata)

view raw JSON →