brickblock 0.1.0

A fun visualisation library for those that like boxes

Sterling: A fun visualisation library for those that like boxes

This is a small library that uses blocks in matplotlib's visually appealing 3D extension - and aims to be the 'seaborn of matplotlib-3D', if you will.

Core abstractions

At the centre of Sterling is the Space. A Space contains objects, and when a user wants a visualisation, they render the current state of the Space - the rendered state is known as a Scene.

Currently there is one object used for composing visualisations in Sterling: the Cube. Cube objects can be added into the Space with a degree of control over their visual presentation: transparency, colour, line widths.

Sterling also supports Transitions and Transforms:

• A Transform is simply any change to the space like adding one or more Cube objects. It does not produce a Scene. These are useful when you are just iteratively building up to a complex Scene.
• A Transition, by contrast, is a transform between two Scenes. These are useful for generating sequences of Scenes - the Space will keep track of its state over time, as well as which states are Scenes. A Transition can be defined by one or more Transforms.

Having these abstractions allows programmers to create animated visualisations like GIFs. You define a Space, using Transforms and Transitions to evolve the state, and the Scene objects are persisted to enable sequences of images for use in GIFs.

import brickblock as bb


# Let's start with adding a few objects to our space!
cuboid_base = np.array((0, 0, 0))
cuboid_dims = np.array((4, 3, 4))
cuboid_opts = {'color': None, 'alpha': 0}
input_cube = bb.Cuboid(base, cuboid_dims, **cuboid_opts)

filter_base = np.array((0, 0, 1))
filter_dims = np.array((3, 3, 3))
filter_opts = {'color': (1., 1., 0.5), 'alpha': 0.3, 'linewidths': 0.3}
filter_cube = bb.Cube(filter_base, filter_dims, **filter_opts)

s = bb.Space()
# Add a 4x3x4 rectangular cuboid to the space.
s = bb.add_transform(s, input_cube)
# Add a 3x3x3 filter to the space with a single cube.
s = bb.add_transform(s, filter_cube)

# Render the state as a scene and use in a plot-like way...
fig, ax = bb.render(s)
# Or take a snapshot to indicate that this is a scene...
state_history = bb.snapshot(s)
# Do more stuff...
# And then when you're ready, generate a GIF-like output...
gif_filename = bb.stream(s, "gif", path_to_file))