PyCharm 2018.3 Help

Using IPython/Jupyter Notebook with PyCharm

Before you start

Prior to executing the tasks of this tutorial, make sure that the following prerequisites are met:

Note that PyCharm automatically installs the dependencies of these packages.

Creating a Jupyter Notebook file

In the Project Tool Window, click Alt+Insert. Then, on the pop-up menu that appears, choose the option Jupyter Notebook and type the file name (here it is MatplotlibExample.ipynb).

The newly created file now shows up in the Project Tool Window and automatically opens for editing .

By now, the new file is empty, but PyCharm recognizes it as a notebook file. As such, this file is marked with the icon Jupyter notebook icon and features a toolbar, which is a complete replica of the real Jupyter Notebook toolbar:

Jupyter notebook toolbar

Filling in and running the first cell

This is most easy. Just click the first cell and start typing. For example, in the very first cell type the following code to configure the matplotlib package:

%matplotlib inline

Next, click the icon run cell icon to run the cell (alternatively, you can press Shift+Enter). PyCharm shows a dialog box, where you have to specify the URL where the Jupyter Notebook server will run:

Jupyter server is running

In this dialog box, click Cancel, and then click the Run Jupyter Notebook link:

Run Jupyter Notebook link

Next, if you didn't install the "Jupyter Notebook" package yet, the run/debug configuration dialog appears showing the error message:

install package inspection

Install the package to fix the problem.

Jupyter server runs in the console:

Jupyter notebook is running in the console

Follow this address:

Notebook is in a browser

Actually, that's it... From now on you are ready to work with the notebook integration.

Working with cells

First of all, add the following import statement:

from pylab import *

This how it's done. To create the next empty cell, click add a cell on the toolbar:

Adding an ampty cell

Start typing in this cell, and notice code completion:

code completion

Click Run a cell on the toolbar again to run this cell. Note that the cell produces no output, but the next empty cell is created automatically. In this new cell, enter the following code:

figure() plot(x, y, 'r') xlabel('x') ylabel('y') title('title') show()

Run this cell. Oops! The attempt to run results in an error:

Error message

It seems that the variables should be defined first. To do that, add a new cell.


Since the new cell is added below the current one, click the cell with import statement - its frame becomes green. Then on the toolbar click Add a cell (or press Alt+Insert) .

In the created cell, type the import statements and run them:

Running with the import fixed

The new cell is created automalically. In this cell, type the following code that will define x and y variables:

x = linspace(0, 5, 10) y = x ** 2

Run this cell, and then run the next one. This time it shows the expected output:

Cell has been executed

Clipboard operations with the cells

You can perform the standard clipboard operations: Ctrl+C, Ctrl+X and Ctrl+V.

Try these shortcuts yourself.

Running and stopping kernels

As you've already learnt, the button Execute a cell is used to execute a cell.

If calculation of a certain cell takes too much time, you can always stop it. To do that, click Stop the execution on the document toolbar.

Finally, you can rerun the kernel by clicking Rerun the kernel on the document toolbar.

The messages about all these actions show up in the console:

Notifications shown in the console

Choosing style

Look at the drop-down list to the right of the document toolbar. It allows you to choose presentation style of a cell. For example, the existing cells are presented as code.

Click the cell with the import statement again, and then click Add a cell. The new cell appears below. By default, its style selector shows Code. In this cell, type the following text:

plot example

Run this cell and see the error message. Next, click the down arrow, and choose Markdown from the list. The cell changes its view:

The cell has changed

Now click Execute a cell on the toolbar, and see how the cell looks now:

Cell is being executed

Now you can just select the desired style from the drop-down list, and the view of the cell changes appropriately:

Alternative style

Writing formulae

Add a new cell. In this cell, choose Markdown from the style selector, and type the following text:

$$c = \sqrt{a^2 + b^2}$$

Click Execute a cell. The result is stunning:

The cell is being executed

As you see, PyCharm's Jupyter Notebook integration makes it possible to use LaTex notation and render formulae, labels and text.

Next, explore the more complicated case. The expected result - the formula - should appear as the result of calculation. Add a cell and type the following code (taken from SymPy: Open Source Symbolic Mathematics):

from __future__ import division from IPython.display import display from sympy.interactive import printing printing.init_printing(use_latex='mathjax') import sympy as sym from sympy import * x, y, z = symbols("x y z") k, m, n = symbols("k m n", integer=True) f, g, h = map(Function, 'fgh')

Run this cell. It gives no output. Next, add another cell and type the following:

Rational(3,2)*pi + exp(I*x) / (x**2 + y)

Click Execute a cell and enjoy:

Final output
Last modified: 18 January 2019

See Also