Setting up: from paper to Python

You can start this book with nothing but a pencil. Each later grade adds one tool. Choose the level that matches your school; every chapter has a version for lower levels.

Level	What you need	Enough for
0. Paper	notebook, pencil, ruler, graph paper, calculator	all of Grade 8; the Track A parts of Grades 9–12
1. Spreadsheet	any computer or phone with LibreOffice Calc, Google Sheets or Excel	Grade 9 Units 1–2, 7; many exercises
2. CODAP	a web browser (internet for first load)	graphing and exploring in Grades 8–9
3. Python in the browser	JupyterLite (works offline once loaded, or from the school bundle)	Grades 9–12 notebooks
4. Python on a computer	Python 3.10+, pandas, matplotlib, scikit-learn, Jupyter	everything, including Streamlit and Git

Level 0: paper

Keep one notebook for the whole year. Draw tables with a ruler; use graph paper for charts. A cheap calculator with square root is enough. The Grade 8 worksheets in curriculum/grade-08/unit-01-what-is-data/worksheets/ can be printed.

Level 1: spreadsheet

• LibreOffice Calc — free, works offline, available in English and Nepali interfaces (https://www.libreoffice.org).
• Google Sheets — free with a Google account; needs internet.
• Microsoft Excel — if your school has it.

All formulas in this book (SUM, AVERAGE, COUNTIF, SUMIF, VLOOKUP, pivot tables) work in all three. To open a CSV file: File → Open, choose the file, and in the import window choose "comma" as the separator and UTF-8 as the character set (so that Nepali text displays correctly).

Level 2: CODAP

CODAP (Common Online Data Analysis Platform, https://codap.concord.org) is a free tool from the Concord Consortium for exploring data by dragging: drop a CSV file onto the page, make a graph, drag a column to an axis. No account is needed. It is ideal for Grade 8 and for anyone who wants to see data before coding.

Level 3: Python in the browser (JupyterLite)

JupyterLite runs Python (through Pyodide) entirely inside your web browser: no installation, no server. The project website hosts a JupyterLite copy of every notebook with the shipped datasets.

1. Open the project website's Lite link (or the lite/ folder of the offline bundle opened with a local web server).
2. Open a notebook from curriculum/grade-XX/unit-…/notebooks/.
3. Press Shift + Enter to run each cell.
4. The first time, the browser downloads Python and pandas (tens of megabytes); after that it is cached.

If a library is missing, run in the first cell: %pip install pandas matplotlib scikit-learn.

Colab (Google's free notebooks) also works with any notebook from the repository, but needs a Google account and a good internet connection.

Level 4: Python on your own computer

1. Install Python 3.10 or newer from https://www.python.org (tick "Add Python to PATH" on Windows), or install Miniconda/Anaconda.
2. Download the curriculum (the GitHub "Download ZIP" button, or git clone the repository).
3. In a terminal, inside the folder:

python -m venv .venv

then activate it (.venv\Scripts\activate on Windows, source .venv/bin/activate on macOS/Linux) and install:

pip install -r requirements.txt jupyterlab

1. Start Jupyter with jupyter lab and open any notebook.
2. For Grade 10 Unit 7 dashboards: pip install -r curriculum/grade-10/unit-07-dashboards-with-streamlit/requirements-streamlit.txt and streamlit run app/air_dashboard.py inside that unit's folder.
3. For Grade 11 Unit 8: install Git from https://git-scm.com.

An old laptop with 4 GB of RAM is enough for everything in this book.

Schools without internet: the offline bundle

On any computer with internet, a teacher can build a school bundle (curriculum, synthetic datasets, licences, and JupyterLite if built) and copy it to USB sticks:

python tools/build_offline_bundle.py

The result, dist/school-bundle.zip, can be copied to every computer in a lab. Real datasets for the exercises can be downloaded once (Appendix A gives the links and licences) and copied the same way, as long as their licences allow it.

Getting data files

• Synthetic practice tables are in datasets/<name>/clean/ — for example datasets/census-districts/clean/districts_synthetic.csv. In this book, code reads them by file name (pd.read_csv("districts_synthetic.csv")): copy the CSV next to your notebook, or use the notebook's data_file() helper, which finds it automatically.
• Real datasets — see Appendix A. Some (World Bank, OWID, USGS, Open-Meteo, NASA POWER) can be read directly from a URL in Python; others (Open Data Nepal, ICIMOD) are downloaded from their web pages.

If something goes wrong

Problem	Likely cause	Fix
FileNotFoundError	the CSV is not where the code looks	put the file next to the notebook or give the full path
ModuleNotFoundError: No module named 'pandas'	library not installed in this environment	pip install pandas (or %pip install pandas in JupyterLite)
Nepali text shows as ??? or boxes	wrong character encoding or font	open/save CSVs as UTF-8; install a Devanagari font such as Noto Sans Devanagari
HTTP Error 403 when reading a URL	the website blocks scripts	add a browser-like user agent (Appendix A, RD-02) or download the file manually
The notebook gives different numbers from the book	different random seed or data version	check random_state/seed; check the data file's date
Everything is slow	very large file or old computer	load fewer columns (usecols=), use smaller types (Grade 11 Unit 1)