Project Overview

Introduction

The visualizations project is a framework written in python for animating Matplotlib based components with data primarily from neural network models.

Motivation

There exists a wide variety of visualization libraries that allow for the creation of moving or animated plots (such as Plotly Dash), however I found that many of these libraries were limited for my particular use case. I needed to display multiple custom plots simultaneously while having the ability to control animation rate, component updates, a various display properties.

The goal of this project is to create a convienent framework for combining and animating matplotlib components concurrently. The product is akin to a dashboard with the single purpose of displaying multiple plots, each with many different sources of sequential data.

Features

Component subplots may be updated on a step or sub-step schedule, meaning that select subplots can display information at different rates or resolutions. The sub-step animation support provides an additional layer of flexibilty when displaying sequential data that may have been collected at different intervals.

See the Examples section for more videos and screenshots generated using this framework.

subnet_plot

This module contains the SubnetPlot and SubnetSubplot template classes that encapsulate the base functionality of this framework. The SubnetPlot class contains and manages child SubnetSubplots. All custom plot components (Bar, Line, Matrix, etc.) inherit from SubnetSubplot.

The Getting Started guide covers the included subplot component.

subnet_visualize

The visualize module contains a collection of combined visualizations for neural networks implemented with the subnet_plot module. This module may be used as an example of ways to combine and animate the base components from subnet_plot.

subnet_animate

The subnet_animate module enables animations to be rendered and saved as .mp4 video files.