•
Five free official tools before you pay for EPAT, CQF, or an agentic bootcamp: CPython, VS Code, Jupyter, pandas, Git. After-work stack for employed HK learners — skip the credential until a .py file runs.
EPAT, CQF, and the 2026 Agentic AI bootcamps are longer credentials. Before you pay that invoice, you need a laptop that can run a .py file, open a table from Excel, and keep last Tuesday's script. This is the foundation stack — not a vendor ranking, and not a list ordered by download volume (we do not have that data).
The 2026 quant-education market is pitching Agentic AI as the enrollment hook. Those products are the longer path. This piece is the stack first. Skip the expensive credential until you can run these five.
All five are free to install from the official sites cited below. Licenses are noted with the source. They do different jobs; this is not a popularity contest.
.py file.py file.py filesIf you only have one sitting, start with Set up Python after work in 30 minutes (Windows, Mac, VS Code). Come back here for why Jupyter, pandas, and Git sit next to that install.
What it is. CPython is the reference Python interpreter hosted by the Python Software Foundation. python.org states that all Python releases are Open Source. Current releases use the PSF License Version 2, which is royalty-free.
Windows (2026 docs). The official Using Python on Windows page (Python 3.14.7 documentation, updated 31 Aug 2026) tells you to obtain Python from the CPython team with the Python Install Manager, downloaded from python.org/downloads or the Microsoft Store. After install, python or py launches a runtime. The traditional full installer is deprecated since 3.14 and will not be produced for Python 3.16 or later — so do not follow a 2023 blog that only mentions python-3.x.x.exe and "Add to PATH" as the whole story. The 2026 path is the install manager, then python myscript.py.
Mac. Use the python.org macOS installer, not the Apple /usr/bin/python3 stub. Latest stable listed there as of this writing: Python 3.14.7 (5 Aug 2026). Official Using Python on macOS (same 3.14.7 docs) recommends the most recent supported version; the package is a universal2 build. After install, run python3 myscript.py in Terminal. Do not delete Apple's system Python.
After-work test. One file, hello.py, that prints a line. If that runs, you have CPython. Syntax can wait for Learn Python after work: 19 short lessons, a few evenings a week.
Price. Free. Official binaries from python.org; license: PSF License Version 2.
What it is. Visual Studio Code is a free editor. Microsoft's product license lets you install any number of copies to develop and test applications. The editor source is published under the MIT license on GitHub; the branded product uses Microsoft's terms. You do not need a paid IDE for this stack.
Python extension (2026). Official page: Python in Visual Studio Code (page dated 19 Aug 2026). Install a Python interpreter separately (the extension is not CPython). Then install the Microsoft Python extension from the Marketplace and use Python: Select Interpreter. The extension runs, debugs, and autocompletes against that interpreter. Microsoft's February 2026 VS Code notes also describe a Python Environments extension rolling out for venv/conda/package UI — useful later, not required tonight.
After-work job. Open a folder, create hello.py, click Run Python File or type python hello.py / python3 hello.py in the terminal. One editor is enough. Do not install a second "just in case."
Price. VS Code is free to download and use. Optional extras (Copilot, some Marketplace extensions) can cost money; none of them are this stack.
.py fileWhat it is. Project Jupyter ships two interfaces: JupyterLab ("the latest web-based interactive development environment for notebooks, code, and data") and classic Jupyter Notebook. Official install from jupyter.org/install:
pip install jupyterlab
jupyter lab
Classic notebook: pip install notebook then jupyter notebook. Homebrew on Mac: brew install jupyterlab. JupyterLab 4.6 is the current 4.x line in the project's own changelog (4.6.0 announced 2 Jul 2026).
Notebook vs .py after work. A .ipynb notebook is a JSON document of cells: code, markdown, and saved output. Good for poking one Excel extract at 21:30 — you see the table under the cell. A .py file is a linear script you run with python file.py. It diffs cleanly, emails cleanly, and is what Git (item 5) versions well. Quant courses and work scripts expect you to ship .py files. Use the notebook to explore; save the logic you keep in a .py.
VS Code can open notebooks too if you add the Jupyter extension; that is optional.
Price. Free via PyPI. Project Jupyter tools are open-source (JupyterLab is BSD 3-Clause on the project's GitHub).
What it is. pandas is "an open source, BSD-licensed library" for tables in Python. Docs at pandas.pydata.org/docs (version 3.0.5, dated 22 Jul 2026). pandas 3.0.0 shipped 21 Jan 2026. One line on the 3.0 change: Copy-on-Write is now the default and only mode — chained assignment will not work the old way. That is a behaviour change, not a reason to skip pandas. For the CoW-vs-Polars evening decision, see Pandas 3.0 Copy-on-Write vs Polars after work. This article does not redo that comparison.
Excel. The official I/O table lists read_excel / to_excel for MS Excel. Install docs: pip install pandas and, for Excel extras, pip install "pandas[excel]". That extra pulls readers such as openpyxl. After-work move: export one sheet, then pd.read_excel("file.xlsx") (or read_csv if you saved CSV). That is the table you already have at work, in Python.
You still need a first program that uses a library, not another print. That is Your first Python program that uses a library (not another print).
Price. Free. pip install pandas from PyPI; BSD-licensed per the pandas docs.
.py filesWhat it is. Git is "a free and open source distributed version control system." Latest source release on the official site: 2.55.0 (29 Jun 2026). Installers: git-scm.com.
Why before a paid course. Course homework and your own .py files will break. Git lets you keep Tuesday's working script when Thursday's edit does not run. You do not need GitHub, a team, or a branching strategy tonight. In the folder where hello.py lives:
git init
git add hello.py
git commit -m "first script that ran"
That is enough to start. Hosting (GitHub, GitLab, a private remote) is a later choice and may require an account; Git itself does not.
Price. Free and open source (official wording on git-scm.com). No purchase for the tool.
A certificate does not install Python for you. If you cannot run a .py file, open a sheet in pandas, and commit the file, an EPAT/CQF/agentic calendar will still start with environment week — except you will have already paid. Stay competitive in the job you have: these five tools are what analysts actually type into after dinner, on a Hong Kong weeknight.
This is not a CEF programme, not a QF qualification, and not a ranking of bootcamps.
When the five tools are boring, learn the language in modest blocks. Introduction to Python Programming is 19 published short lessons. Check the course page for the current enrollment window. That is the only course this article points you to enroll in.
read_excel): pandas.pydata.org/docs/user_guide/io.html