matviz

1.0.6 verified Fri May 01 auth: no python

matviz provides simple matrix data visualization tools built on matplotlib and numpy. Current version 1.0.6 supports Python >=3.10. Development appears intermittent with no recent releases since v1.0.0 (July 2023).

pip install matviz

Common errors

error AttributeError: module 'matviz' has no attribute 'plot_matrix' ↓

cause Using 'import matviz' and then 'matviz.plot_matrix()' instead of direct import.

fix

Use 'from matviz import plot_matrix' and then 'plot_matrix(data)'.

error ValueError: could not broadcast input array from shape (n,m) into shape (n,) ↓

cause Passing a 2D array but with incompatible dimensions for the selected color mapping or axis.

fix

Ensure input is a 2D numpy array of shape (rows, cols). Use np.array(data).reshape(rows, cols) if needed.

Warnings

gotcha plot_matrix returns a figure and axes tuple, not just a figure. Unpack with fig, ax = plot_matrix(...) to avoid AttributeError. ↓

fix Use fig, ax = plot_matrix(data) instead of fig = plot_matrix(data).

gotcha The library relies on matplotlib's interactive backend. If running in a headless environment, set matplotlib backend to 'Agg' before importing matviz. ↓

fix import matplotlib; matplotlib.use('Agg') before import matviz.

deprecated Functions like heatmap and correlation_plot from early versions (pre-1.0.0) are deprecated. Use plot_matrix instead. ↓

fix Upgrade to 1.0.0+ and use plot_matrix.

Imports

plot_matrix
wrong
```
import matviz
```
correct
```
from matviz import plot_matrix
```
Direct import of plot_matrix is required; the module does not expose functions via dot notation.
MatrixPlotter
wrong
```
from matviz.plotter import MatrixPlotter
```
correct
```
from matviz import MatrixPlotter
```
MatrixPlotter is directly exposed at package level, not in a submodule.

Quickstart

Create a basic matrix heatmap with matviz.

import numpy as np
from matviz import plot_matrix

data = np.random.rand(5,5)
fig, ax = plot_matrix(data, title='My Matrix')
import matplotlib.pyplot as plt
plt.show()