This package allows for simple visualisation of data collected (e.g. from machine learning experiments). It is primarily designed for complimentary use with this data logger package and this runner package.

This package is written in Python 3. Requirements are minimal:

• matplotlib
• pandas
• numpy

Installation can either be performed by cloning the repository and running pip install -e . from the package root, or via pip install -e git://github.com/seblee97/plotter.git#egg=plotter-seblee97.

The purpose of the package is to facilitate visualisation of ablations in machine learning experiments. The Plotter class in plotter.py is for creating plots of a range of scalar values logged during a specific experiment. Other functions in plot_functions.py allow for visualising and comparing different runs including consolidating seeds and showing variation across seeds.