Notebook Labs - Noro-Official/OQS GitHub Wiki

Notebook Labs and Assignments

This page provides examples, design tips, and ready-to-use structures for building Jupyter-based labs and assignments with OpenQStack.

These materials are ideal for:

Undergraduate quantum computing courses
Graduate-level QEC modules
Remote learning environments (Colab, Binder)
Project-based learning or open exploration

Why Notebooks?

Jupyter notebooks let students:

Write and run real QEC code
Visualize states, errors, and recovery processes
Document their thought process alongside execution
Get immediate, visual feedback on correctness

This bridges theoretical knowledge with intuitive understanding.

Example Lab: Bit-Flip Recovery

Title: Protecting |+⟩ with a Bit-Flip Code

from openqstack.qec import BitFlipCode
from openqstack.visualize import show_state, plot_bloch
import numpy as np

psi = [1/np.sqrt(2), 1/np.sqrt(2)]
code = BitFlipCode()

encoded = code.encode(psi)
corrupted = code.apply_random_X_error(encoded)
syndrome = code.measure_syndrome(corrupted)
recovered = code.recover(corrupted, syndrome)
decoded = code.decode(recovered)

plot_bloch(decoded)

Learning goals:

Understand syndrome-based error detection
Visualize logical state restoration
Reflect on limits of the code (e.g. double errors)

Notebook Design Tips

Start with scaffolded code: have students fill in a line or two, not 50
Include visual checkpoints with show_state() or plot_probabilities()
Use assert np.allclose(...) to validate answers in real time
Add markdown cells that guide thinking, not just procedures

Assignment Ideas

1. Write Your Own Error Channel

class PhaseDampingChannel(ErrorChannel):
    def kraus_operators(self):
        ...

Evaluate:

Correctness of Kraus representation
Physical behavior under test states
Ability to describe limitations

2. Analyze a Faulty Decoder

Give students:

A hard-coded syndrome-to-correction map
Cases where it fails
Ask them to propose fixes or build better logic

3. Logical Error Rate Under Noise

Loop over multiple runs:

success = 0
for _ in range(1000):
    encoded = code.encode(psi)
    noisy = channel(encoded)
    recovered = code.recover(noisy, code.measure_syndrome(noisy))
    final = code.decode(recovered)
    if np.allclose(np.abs(final), np.abs(psi)):
        success += 1

Let students:

Graph logical success vs. noise probability
Identify phase transitions
Compare bit-flip vs. depolarizing

Running Environments

Local

pip install jupyterlab
jupyter lab

Google Colab

!git clone https://github.com/OpenQStack/OQS.git
%cd openqstack
!pip install -e .

Then open a notebook and begin importing OpenQStack.

Coming Soon

notebooks/ directory in the repo
Prebuilt .ipynb labs for classroom drop-in use
Binder links to run notebooks in-browser