Quick answer: Jupyter Notebook is the working environment where most data analysis happens: code, output, notes and charts in one document, executed in cells. Launch it from an activated environment with jupyter notebook (or use VS Code’s built-in notebook support). The essential workflow: Shift+Enter runs a cell and moves down, and each notebook keeps its variables in memory between cells. This guide covers the launch, the five shortcuts that matter, and the habits that keep notebooks clean.
Launching Jupyter
conda activate myenv # or: venv\Scripts\activate
jupyter notebook # opens in your browser at localhost:8888If you installed Anaconda, Jupyter is included and launchable from Anaconda Navigator too. The browser tab is the interface; the kernel (Python process) runs behind it. Closing the browser tab does not stop the kernel — shut it down from the Jupyter dashboard or terminal (Ctrl+C twice).
The five shortcuts that matter
| Shortcut | Mode | Does |
|---|---|---|
Shift+Enter | either | Run cell, advance to next |
A / B | command | New cell Above / Below |
DD (twice) | command | Delete cell |
M / Y | command | Cell becomes Markdown / Code |
Esc / Enter | any | Switch command ↔ edit mode |
Command mode (cell outlined blue, no cursor) is for structure; edit mode (green, cursor) is for typing. Learning this distinction turns Jupyter from confusing to fast within a day.
How cells actually work
All cells share one Python process — import pandas in cell one, use pd everywhere after. That power is also the danger: run cells out of order and your variables no longer match the notebook’s story. The professional habits:
- Restart & Run All before sharing (Kernel menu) — proves the notebook works top-to-bottom.
- One idea per cell — a load, a filter, a plot; outputs stay readable.
- Markdown cells for narrative — press
Mand write why, not just what. Future-you is the audience. - Watch the number in
In [n]— it shows execution order; out-of-order numbers reveal a stale notebook.
A first complete session
# Cell 1 - setup
import pandas as pd
df = pd.read_csv('students.csv')
# Cell 2 - look
df.head()
# Cell 3 - one real question
df.groupby('city')['score'].mean().sort_values(ascending=False)Three cells, one import, one inspection, one answer — that is the shape of most exploratory analysis. Combined with the pandas operations from our first DataFrame guide, you have a working analytics setup on day one.
For instructor-led practice with notebooks, datasets and review, Ampersand Academy teaches Python analytics one-to-one.
Frequently asked questions
How do I start Jupyter Notebook after installing Anaconda?
Open Anaconda Navigator and click Launch under Jupyter Notebook, or open a terminal with your environment active and run jupyter notebook. It opens in your default browser on localhost:8888.
What is the shortcut to run a cell in Jupyter?
Shift+Enter runs the current cell and selects the next one. Ctrl+Enter runs it in place. Alt+Enter runs it and inserts a new cell below.
Why do my variables disappear when I reopen a notebook?
The kernel keeps variables in memory only while running. Reopening the file does not re-execute anything – use Kernel then Restart & Run All to rebuild state from the top.
Jupyter Notebook or JupyterLab – which should I use?
JupyterLab is the modern interface with tabs, a file browser and side panels. Classic Notebook is simpler. Both run the same kernels; start with JupyterLab.
Can I use Jupyter in VS Code instead?
Yes – install the Python and Jupyter extensions, open a .ipynb file, and select your kernel. Many professionals prefer VS Code notebooks for the integrated editor and git workflow.