Jupyter Shortcuts & Magic Commands

Why shortcuts and magics matter

In data analytics, you often:

• Rerun code many times
• Explore different plots and transformations
• Measure performance
• Run shell commands

Jupyter shortcuts and magic commands help you iterate much faster.

Keyboard shortcuts (classic Jupyter Notebook)

Jupyter has two main modes:

• Command mode (press Esc): navigate and manage cells
• Edit mode (press Enter): edit inside a cell

Most useful shortcuts (Command mode)

• A → insert cell above
• B → insert cell below
• D, D → delete selected cell
• X → cut cell
• C → copy cell
• V → paste cell
• Z → undo delete
• M → convert cell to Markdown
• Y → convert cell to Code
• Shift + Up/Down → select multiple cells

Most useful shortcuts (Edit mode)

• Ctrl + Enter → run cell (stay)
• Shift + Enter → run cell (move down)
• Alt + Enter → run cell (insert new cell below)
• Tab → autocomplete (when available)
• Ctrl + / → toggle comment in selection

Tip

Shortcut cheat sheet In classic notebook: Help → Keyboard Shortcuts.

What are magic commands?

Magic commands are special Jupyter commands that start with:

• % for line magics
• %% for cell magics

They are not standard Python syntax; they’re handled by IPython/Jupyter.

Common line magics

%time and %timeit

Use these to measure execution time.

time

%time sum(range(10_000_00))

%timeit runs the expression multiple times and gives an average:

timeit

%timeit sum(range(1_000_00))

%who and %whos

Show variables currently defined.

who

%whos

%pwd and %cd

• %pwd shows current working directory
• %cd changes it

pwd

%pwd

cd

%cd data

%ls (list files)

ls

%ls

Plotting magics

%matplotlib inline

Classic way to render plots inside the notebook.

matplotlib-inline

%matplotlib inline

%matplotlib notebook (interactive)

This can enable interactive plots in the classic notebook.

matplotlib-notebook

%matplotlib notebook

Note

In newer setups, Matplotlib interactivity depends on your environment and extensions. If %matplotlib notebook doesn’t work, you can still use inline plots.

Useful cell magics

%%writefile (save code to a file)

You can write a cell’s content into a Python script.

writefile

%%writefile my_script.py
print("Hello from a file")

%%bash (run bash commands)

If your Jupyter supports it, you can run bash scripts:

bash

%%bash
ls -la

If %%bash isn’t available, you can use ! shell escapes instead.

Shell escapes with !

Prefix a command with ! to run it in the OS shell.

shell-escape

!python --version

pip-list

!pip list | head

Getting help quickly

Add ? to functions

help

import pandas as pd
pd.DataFrame?

Use help()

help2

help(pd.DataFrame)

Visualize it

The % prefix tells IPython how much of the input to treat as a magic command versus plain Python.

diagram Magic command dispatch mermaid

  flowchart TD
  A["Line typed in a cell"] --> B{"Starts with?"}
  B -->|"%%"| C["Cell magic: applies to the WHOLE cell"]
  B -->|"%"| D["Line magic: applies to ONE line"]
  B -->|"!"| E["Shell escape: runs in the OS shell"]
  B -->|"nothing special"| F["Normal Python code"]

How IPython decides between a line magic, a cell magic, and a shell escape

From the book

McKinney notes that IPython’s numbered In [n]: prompt (instead of the plain >>>) is what makes features like %run and tab completion possible — IPython is tracking your whole session, not just one line at a time. He also highlights ? as a fast way to pull up a function’s docstring (e.g. pd.DataFrame?) without leaving your workflow to search documentation online.

Common mistakes

Mistake 1: Using magics in .py files

Magics only work in Jupyter/IPython, not in normal Python scripts.

Mistake 2: Forgetting to rerun earlier cells

Notebook state is incremental. If you restart the kernel, rerun setup cells.

Next

Continue to: Google Colab Walkthrough to learn cloud notebooks and GPU/TPU options.

🧪 Try It Yourself

Exercise 1 – Classify a Magic Command

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Time a Function With timeit

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Strip the ! From a Shell Escape

pch.exerciseTag pch.exerciseTitle