Pre-course setup — due before Session 1

The program prep session installs the MSBA-standard stack — Python 3.13, uv, VS Code, Git, and a GitHub account — for every course, this one included. If you attended it, skip to §2 Run the check (everyone runs the check). §1 is the same installs, for anyone who missed the session or is on a fresh machine.

Session 1 cannot be tech support. Submit the check’s output to the Moodle “Setup verification” slot before the first class. If anything fails, come to the “uv install troubleshooting” office-hours slot in the week before Session 1 — with the output.

1. Install the tools (skip if you did the prep session)

Git (2.23 or later)

macOS: open Terminal, type git --version. If it offers to install the command-line tools, accept. Or brew install git.
Windows: you need an administrator account for the installs below (“Access is denied” partway through means you are on a Standard account). Install Git for Windows. Accept the defaults, with two exceptions: on the default editor screen choose Visual Studio Code if it is offered (it is greyed out until VS Code is installed — then pick anything and, after installing VS Code below, run git config --global core.editor "code --wait"); on the initial branch screen tick Override the default branch name and enter main. This installs Git Bash, which you will use for everything in this course — not PowerShell, not cmd, not Anaconda Prompt. Then, in Git Bash: git config --global core.autocrlf input
Linux: sudo apt install git (or your distribution’s equivalent).

uv (the program-standard Python manager)

macOS / Linux: in Terminal: curl -LsSf https://astral.sh/uv/install.sh | sh — then close and reopen the terminal.
Windows: in Git Bash the same curl line works. If it does not, open PowerShell once for powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex", then go back to Git Bash. Close and reopen Git Bash.

Check: uv --version.

Python (3.13 — the program standard)

You do not need to install Python separately: every course repo pins 3.13 in its .python-version, and uv sync downloads it automatically the first time. If you want it on hand anyway: uv python install 3.13.

Windows: if python -c "import sys; print(sys.executable)" prints a path containing WindowsApps, that is the Microsoft Store stub, not Python. Disable it under Settings → Apps → Advanced app settings → App execution aliases (turn off both python entries), then re-check.

VS Code (the program-standard editor)

Install VS Code and, inside it, the Python and Jupyter extensions (Microsoft).

Windows — set the integrated terminal to Git Bash (VS Code’s default is PowerShell, which this course does not support): open the Command Palette (Ctrl+Shift+P) → Terminal: Select Default Profile → Git Bash. Every terminal you open inside VS Code is then the same shell the course teaches.

Notebook standard

Every course notebook runs the same way, and the setup check (§2) makes you do it once:

Open the project folder in VS Code (File → Open Folder…), never the notebook file on its own. If VS Code asks whether you trust the authors, choose Yes, I trust the authors — in Restricted Mode the kernel picker shows no environments at all.
Open the notebook and click Select Kernel (top right) → Python Environments… → the entry marked Recommended whose path contains .venv. That is the environment uv sync built for this project.
Verify with a cell: import sys; sys.executable must print a path containing .venv.

You do not need to install anything notebook-related globally: each course repo brings its own notebook kernel, and VS Code uses it once you open the project folder and pick the .venv interpreter. Verify with import sys; print(sys.executable) in a cell — the printed path is the proof, not the picker’s list. Do not install Anaconda for this course; if you already have it, that is fine — just never select an Anaconda kernel for course notebooks.

GitHub account

Not required. Starter repos are public; you clone them, you never push. All submissions go through Moodle.

2. Run the check

In your terminal (macOS Terminal / Git Bash):

git clone https://github.com/earino/ecbs5293-setup-check.git
cd ecbs5293-setup-check
uv sync
uv run python check.py

It prints a short report ending in ALL CHECKS PASSED or FIX THESE FIRST with one line per problem. Fix, re-run.

Then the half a script cannot check: open the cloned folder in VS Code, open check_notebook.ipynb, pick the .venv kernel (the Notebook standard above), and Run All. Three outputs appear: a path containing .venv, a pandas version, and the folder. Submit both the report and a screenshot of those three outputs to the Moodle slot.

What it proves: uv sync installed this project’s packages into an environment of its own; uv run found that environment’s Python; that Python can import pandas; and VS Code can execute a cell on that environment, which is exactly what Lab 1 asks of you in its first five minutes. Session 2 explains each step — for now, do it exactly.

3. Rules that save you an hour

Windows: never type bare python in Git Bash. On some setups it hangs with no error; on others it drops you into Python’s >>> prompt. Do not find out which you have: always python --version, python -c "...", python script.py, or uv run python ….
Notebooks run on the project’s own kernel. In VS Code: open the folder, pick the .venv/ interpreter in the kernel picker, and verify with import sys; sys.executable. A global Python, Anaconda, or another project’s kernel silently uses a different Python than the one uv manages — and recreates exactly the problem this course teaches you to diagnose.
Every course repo is run from its own folder (the one containing pyproject.toml). The READMEs say so; the labs are built around it.

4. If it still does not work

Submit the failing output anyway — that is useful — and book the office-hours slot. The most common fixes are: reopening the terminal after installing uv; the Store stub on Windows; a corporate proxy blocking the installer (use a personal network once); and an old Anaconda install shadowing python on PATH.