Setting up your coding environment - CPSC 330 Applied Machine Learning 2026W1

Working in PrairieLearn¶

You can complete assignments in the PrairieLearn workspace, which provides the coding environment in your browser. Follow the workspace instructions in each assignment. No local installation is needed for this workflow. You can also read the course book in your browser.

The rest of this guide is optional: use it if you want to run lecture notebooks and demos or work on assignments on your own computer. It assumes you are comfortable opening a terminal and navigating between folders.

Local setup: install the tools once¶

We use uv to manage Python and packages, and JupyterLab or VS Code to run notebooks. The lecture environment files live in the cpsc330-book repository. The cpsc330-2026W1 repository contains course logistics and these instructions.

Use PowerShell on Windows and your usual Terminal on macOS or Linux. After installing uv, the commands below are the same on all three platforms. You do not need WSL, Conda, or a separate Python installation for these instructions.

Install uv¶

Use the command for your operating system from the official uv installation guide.

Windows (PowerShell):

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

macOS or Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Close and reopen your terminal after installation, then check:

uv --version

You should see something like this:

uv 0.12.5 (210d1f678 2026-08-14 aarch64-apple-darwin)

Your version number, build date, and platform details may differ. Seeing a uv version confirms that the command is available in your terminal.

Install Git and choose an editor¶

Install Git if you want to clone and update the book repository or use Git for assignment work. Git is optional if you only work from assignment ZIP files.

JupyterLab: installed with the course environment; no separate installation step is needed.
VS Code: install VS Code and Microsoft’s Python and Jupyter extensions.

Run lecture notebooks and demos locally¶

In a terminal, navigate to the folder where you want to keep the book, then run these commands one line at a time:

git clone https://github.com/UBC-CS/cpsc330-book.git
cd cpsc330-book
uv sync --locked

uv uses the supplied .python-version file to select Python 3.12, downloading it if needed, and installs the packages into a .venv folder. The first installation may take a while because the course includes large machine-learning libraries.

The pyproject.toml file specifies the dependencies, and uv.lock records their resolved versions. --locked checks that these files agree without changing the lockfile. Keep both files as supplied by the course.

Choose one of the following ways to open your notebooks. These steps also apply to an assignment folder after you have created its environment.

Option A: JupyterLab¶

From the folder containing pyproject.toml and uv.lock, run:

uv run --locked jupyter lab

Open a notebook in the browser window that appears. Choose Python 3 (ipykernel) if prompted for a kernel. Keep the terminal running while you work; when finished, save your notebooks and press Ctrl+C in the terminal to stop the server, confirming shutdown if prompted.

Use this launch command each time so JupyterLab runs in that folder’s environment. No manual environment activation is required.

Option B: VS Code¶

In VS Code, choose File → Open Folder and open cpsc330-book (or the extracted assignment folder).
Open a notebook and click Select Kernel at the top right. Choose Select Another Kernel, if shown, then Python Environments and the .venv belonging to this folder.
Run cells with Shift+Enter.

If you also work with Python scripts, use Python: Select Interpreter in the Command Palette to select the same environment. Notebook kernel selection is a separate step.

The interpreter paths are:

Platform	Interpreter inside the project folder
Windows	`.venv\Scripts\python.exe`
macOS / Linux	`.venv/bin/python`

If the environment does not appear, make sure uv sync --locked completed successfully, then reload the VS Code window. See uv’s Jupyter and VS Code guide for more details.

Update the lecture materials¶

From the cpsc330-book folder, run:

git pull
uv sync --locked

Restart any running notebook kernels after updating the environment. Save personal copies of notebooks before editing them; edits to tracked course files can conflict with later updates. If Git reports a conflict, preserve your work and ask for help before discarding changes.

Work on assignments locally¶

PrairieLearn remains available if you prefer to work in the browser. For local work, use the assignment’s downloadable ZIP and its accompanying instructions.

Extract the entire ZIP into a folder of your choice, outside the cloned book and logistics repositories. Keep the notebook and data in their supplied relative locations.
Check that the extracted folder contains pyproject.toml, uv.lock, and .python-version, alongside the notebook and data. These files specify the environment for that assignment. If they are missing, check the assignment instructions or ask the teaching team for the local setup files.
Open a terminal in that folder and run:
```
uv sync --locked
```
Launch JupyterLab with uv run --locked jupyter lab, or open the assignment folder in VS Code and select that folder’s .venv, as described above.
Save your work and follow the assignment’s instructions to upload the required files to PrairieLearn and submit. Files saved on your computer do not automatically appear in PrairieLearn.

Each assignment has its own environment. Use its supplied environment files even if you have already set up the book environment. You do not need to clone either course repository to use an assignment ZIP that includes these files.

Optional: use a private Git repository¶

You can turn the extracted assignment folder into a private Git repository. Where collaboration is permitted by the assignment, share it only with your authorized partners. Keep your notebook and the supplied environment files in version control. Add .venv/ and .ipynb_checkpoints/ to .gitignore; follow the assignment’s guidance about tracking data files.

Each collaborator clones the private repository, runs uv sync --locked, and selects their own local .venv. Do not commit or share the .venv folder itself.

Check your environment¶

Run this cell in a notebook opened through JupyterLab or VS Code:

import sys
import numpy as np
import pandas as pd
import sklearn
import matplotlib.pyplot as plt

print("Python:", sys.version)
print("Interpreter:", sys.executable)
print("scikit-learn:", sklearn.__version__)
plt.plot([0, 1, 2], [0, 1, 4])
plt.show()

You should see version information and a small plot. The interpreter path should point inside the .venv of the book or assignment folder you are working in.

Troubleshooting¶

uv is not recognized: restart your terminal (and VS Code if using its terminal). If the problem persists, consult the uv installation guide.
No pyproject.toml found: navigate to the book root or the extracted assignment folder containing the environment files, then rerun the command.
A notebook cannot import a package: check sys.executable using the cell above. Select the correct kernel, run uv sync --locked in the corresponding folder, and restart the kernel.
A data file cannot be found: extract the complete assignment ZIP and preserve its folder structure. For lecture notebooks, follow any data-download instructions in the notebook.
The lockfile is out of date or installation fails: for the book, first obtain the latest course files with git pull. For assignments, use the files supplied with that assignment. Share the error message, operating system, and uv --version with the teaching team. Avoid removing dependencies or upgrading packages to work around the error, since that changes the course environment.

Some demos may require additional software or downloaded models; follow the instructions provided with those demos. If local setup prevents you from working on an assignment, use its PrairieLearn workspace and bring the error to office hours or tutorials.