Introduction to Workspace objects

This tutorial briefly explains how pyGSTi visualizes data using Workspace objects. Understanding the basics covered here will help demystify the creation of plots and tables in other tutorials.

PyGSTi prefers to use HTML-based figures (native HTML tables and Plotly plots) over the perhaps more traditional LaTeX tables and matplotlib plots. There are several reasons for this:

  1. Interactivity - HTML allows plots and tables to be interactive; with LaTeX this is impossible and with matplotlib it's painful.
  2. HTML's ability to be integrated into web pages (making nicer reports than a many-page PDF) and into Jupyter notebooks.
  3. Portability - Plotly figures (HTML and JS) can be more robustly stored and transported (e.g. over the web) than matplotlib Figure objects, which are difficult even to pickle with Python.

The creation of (HTML) figures, both tables and plots, is handled by the pygsti.report.Workspace factory object.

In [ ]:
import pygsti
ws = pygsti.report.Workspace()

Within an IPython notebook like this one, we can create figures in notebook cells by calling (once, usually at the beginning of the notebook):

In [ ]:
ws.init_notebook_mode(autodisplay=True)

This injects necessary HTML and JavaScript into the notebook so that plots and tables display properly. If everything works properly, you'll see a GREEN "Notebook Initialization Complete" message. If instead you see a BLUE "Loading..." message, then 1) check that this notebook is "Trusted" in the upper right corner of this window and 2) check that you have a working internet connection. You will need to reload this notebook using your browser's reload button after fixing either of these issues.

Setting autodisplay=True means that figures will be displayed as soon as they're created (otherwise we'd have to capture the returned object and call .display() on it). By typing ws. and then hitting TAB you can see the somewhat-descriptive names of the figures that can be created. Here are a few examples (for more, see the Workspace examples tutorial):

In [ ]:
import numpy as np
ws.MatrixPlot( np.array([[1,2],[3,4]],'d'), color_min=0, color_max=4 )
In [ ]:
from pygsti.modelpacks import smq1Q_XYI
ws.GatesTable( smq1Q_XYI.target_model() )

Thats covers the basics! The Workspace examples tutorial shows a gallery of many of the tables and plots a Workspace can create, and the Workspace switchboard tutorial shows how to integrate workspace figures with switches (dropdown boxes, buttons, and sliders). Workspace objects are used internally when generating HTML reports. The report generation tutorial demonstrates how the automated use of a Workspace can lead to a standalone HTML report.