Jupyter notebooks are .ipynb files, where Python code is broken up into cells. Cells can be run, edited and rerun, in whatever order you like. You can see the result of a cell immediately below it.
In addition, you can include Markdown formatted text throughout the document, to provide headings, descriptions, notes and links.
They are a convenient way to write Python code that:
- has a linear flow
- has multiple sections/steps
- requires text explanation
- returns images or plots
- you want to share with others and allow them to reproduce your work
They are often used in data science, since you often follow these steps:
- load some data
- clean it
- find information about it
- plot some results
- export it
They are a useful teaching tool because it allows you to:
- break code down into small chunks (in cells)
- cells can be edited, and rerun by the learner
- formatted text can provide a lot more information and is more readable than comments
Jupyter Notebook is the original way of working with notebooks.
When you choose to launch a notebook, it opens up in a new browser tab.
It has more limited capabilities.
Jupyter Lab is more like an IDE in the browser.
It provides a directory structure on the left, like PyCharm. And when you open a file, it opens in tabs on the right.
One great feature is the Table of contents view that gives you an overview of the headings in the document and allows you to navigate to different sections quickly.
Select a code cell and click the run button to run it (or press SHIFT + ENTER).
This restarts the Python session, clearing any variables, and runs the whole notebook from top to bottom.
This adds a cell below the currently selected one.
- Code: Python code
- Markdown: formatted text (guide)
- Raw: raw text
Note: You must run Markdown cells to see the formatted version
Hover over the gutter beside a cell and click + drag.
Note: In PyCharm Professional, you must select a cell and click the up or down arrows to change its position.