Integrated Development Environments (IDEs)

Integrated Development Environments (IDEs)#

The teaching contents for programming in this eBook require so-called Application Programming Interfaces APIs) and Integrated Development Environments (IDEs).

An API represents a computing interface that enables interactions between multiple software intermediaries. Modular programming becomes easy with an API because it systematically hides complex information that is not necessarily needed to write code according to industry standards. For instance, an API can define the interface between an application (such as Python or Word) and an Operating System (OS) such as Windows, Linux, or macOS (also referred to as platform).

An IDE enables the definition of a project to use, for example, a specific Python environment, and it enables robust coding by pointing out issues directly in the code, even before it runs for the first time. Powerful IDEs go even further and assist in documenting code with markdown (.md files) and directly pipe into git (see the Version Control : git).

Which IDE to choose?

The answer to this question depends on the platform you are using (e.g., Windows or Linux), your personal preferences, and your goals.

For writing Python software itself, PyCharm is a powerful solution. In addition, Jupyter is a great tool for writing word-office-like documents with functional code examples. To test and run Python code (software) locally, for Windows users, the installation of Anaconda is almost indispensable. Linux users will be mostly fine with their system setup without the need to install Anaconda.

For code documentation, examples, and the best learning experience in the Python courses featured in this eBook, consider installing JupyterLab locally. Windows users find instructions in the Jupyter on Windows section. Linux users find instructions in the Jupyter on Linux section.

Once you have an IDE installed, carefully read the instructions for installing Python.

Anaconda#

Anaconda is a powerful tool for managing Python environments on Windows. Linux users better use virtual environments (read more in the chapter on installing Python).

Anaconda Navigator#

Anaconda is a Python and R distribution that enables the usage of a couple of IDEs such as PyCharm, Spyder, or JupyterLab (Notebook).

The very first step to getting started with Anaconda consists in downloading and installing Anaconda where students may use the individual license for educational training purposes (note that a commercial license needs to be purchased for for-profit organizations). On Windows, Anaconda should be installed in the LOCAL user folder (e.g., C:\users<your-user-name>\AppData\Local). Linux or macOS users find download and installation instructions directly at the developer’s website, tailored for their specific distribution, even though they might be better of with virtual environments.

After the successful installation of Anaconda, IDEs for Python programming or markdown editing can be directly installed by launching the Anaconda navigator. conda environments can be created later. Learn more about installing Anaconda (with Python) and this eBook’s support package called flusstools in the Python conda quick guide section and in the video below.

Python Anaconda Installation Video on YouTube

Sebastian Schwindt@hydroinformatics on YouTube.

Miniconda#

Anaconda may cause large environments that require several gigabytes of storage. To install lightweight environments, use Miniconda. Miniconda does not include Anaconda Navigator and to enable working with Jupyter notebooks (in Windows):

  1. Click on Start.

  2. Type Anaconda Prompt and hit enter (use Miniconda3). A Terminal window (black background) opens.

  3. In Anaconda prompt type conda install jupyter and confirm with y when the Terminal asks Proceed ([y]/n)?.

To work with Jupyter notebooks (open, create, or modify), type jupyter lab (or jupyter notebook) in Anaconda Prompt (Miniconda3) and hit Enter. The JupyterLab application will open in the default web browser.

PyCharm#

Jetbrains PyCharm is a powerful but proprietary IDE. Its usage is still free for non-commercial use in education. Alternatives are Spyder IDE (for Python) or RStudio (R and Python). However, before launching any project in an IDE, the installation of an interpreter (e.g., Python or R) is necessary (see chapter on Python installation).

Get PyCharm from the developer’s website or use it through Anaconda. For the educative training purposes provided in this eBook, you may be eligible to use the free education license. To use PyCharm with Anaconda, visit https://docs.anaconda.com.

JupyterLab#

Jupyter is a spin-off of IPython, which is “a rich architecture for interactive computing”. JupyterLab is a product of the nonprofit organization Project Jupyter, which develops “open-source software, open-standards, and services for interactive computing across dozens of programming languages”. A Jupyter notebook (.ipynb file) enables the combination of markdown text blocks with executable code blocks. Essentially, a Jupyter Notebook is a JavaScript Object Notation (JSON file. The structure of JSON files enables the easy export of .ipynb notebooks to many other open standard output formats such as HTML, LaTeX, Markdown, Python, presentation slides, or PDF. The Jupyter kernels support the three core programming languages Julia, Python and R, and many more Jupyter kernels (currently 49) for other programming languages exist.

Working with Jupyter

Get familiar with JupyterLab, by creating files, adding new Markdown or Python cells, and running cells. The essentials of markdown are explained in the Markdown section (short read). Learning Python is more than a short read and the Python Basics chapter provides some insights (takes time).

Jupyter on Windows#

Anaconda Navigator alternatively provides the application Jupyter Notebook. However, JupyterLab is the Project Jupyter’s next-generation user interface, which is more flexible and powerful. This is why this website refers to JupyterLab rather than the Jupyter Notebook app. The following sections explain how to install it on your Windows computer, either by using the graphical user interface of Anaconda Navigator, or the conda prompt command line (recommended).

Via Anaconda Navigator#

  1. Open Anaconda Navigator and make sure to be in the Home tab.

  2. Look for JupyterLab and click on the Install button (if already installed, there is only a Launch button visible).

  3. After successful installation, open JupyterLab, by clicking on the Launch button.

  4. JupyterLab opens in the default web browser, where Jupyter notebooks (.ipynb) or Python files can be created and edited.

Extensions and Spellchecker#

Many additional features for JupyterLab are available through nbextensions, which can be installed through Anaconda Prompt:

conda install -c conda-forge jupyter_contrib_nbextensions

When reading through the Python tutorials on this website, you will probably find one or another spelling mistake (please report mistakes!). In particular, the Python sections may be affected because they were created with JupyterLab, where there is no spell checker pre-installed. To avoid at least the most unpleasant errors you can install a spellchecker in Jupyter. One solution is to install @ijmbbarrs spellchecker, which requires installing nodejs (through Anaconda Prompt and in addition to nbextensions):

conda install -c conda-forge nodejs
jupyter labextension install @ijmbarr/jupyterlab_spellchecker

The spellchecker uses Typo.js as a dictionary and only identifies misspelled words without proposing corrections. More details on spellchecking are available at the developer’s website.

In the case that several warning messages occur when starting JupyterLab (such as [W 18:49:22.283 NotebookApp] Config option template_path not recognized by LenvsHTMLExporter. Did you mean one of: template_file, template_name, template_paths?), downgrade Jupyter Notebook from version 6.x to 5.6.1 (there is currently an issue with the temp_path variable):

conda install "nbconvert=5.6.1"

Jupyter on Linux#

To install JupyterLab on Linux, open Terminal and make sure that pip/pip3 is installed:

sudo apt install python3 python3-pip python3-venv

Export the user-level bin to the PATH environment and install JupyterLab in the user space with the following commands:

export PATH="$HOME/.local/bin:$PATH"
pip install --user jupyterlab

Note

It might be necessary to replace pip with pip3 (depending on the Linux distribution).

To start JupyterLab tap:

jupyter-lab

The command jupyter-lab starts a localhost server that runs JupyterLab, which will open up in a web browser like an interactive website.

Warning

Closing Terminal will also terminate the localhost that runs JupyterLab. Thus, do not close Terminal as long as you are working with JupyterLab, in particular, when there are unsaved books.

A Debugger for Jupyter#

To better understand and troubleshoot code crashes, a debugger represents a great relief. Unfortunately, debugging in Jupyter can cause some headache in the absence of an inherent debugging tool. To get a debugger working with Jupyter, check out this blog entry from Jupyter Project.

Sublime#

Sublime is one of the most popular editors for multiple (computer) languages. However, it is commercial software that is only free to use during an evaluation period without time limit. Read more about it at sublimetext.com.

To install it on Debian Linux platforms, open Terminal and tap (source: https://www.sublimetext.com/docs):

wget -qO - https://download.sublimetext.com/sublimehq-pub.gpg | gpg --dearmor | sudo tee /etc/apt/trusted.gpg.d/sublimehq-archive.gpg

Then select the stable channel (the dev channel has more features but also more bugs):

echo "deb https://download.sublimetext.com/ apt/stable/" | sudo tee /etc/apt/sources.list.d/sublime-text.list

Finally, update apt and install Sublime:

sudo apt update
sudo apt install sublime-text

If an error message occurs, make sure apt works with https sources:

sudo apt install apt-transport-https

When working with sublime, consider using an advanced spell check package, such as LanguageTool. More useful packages for sublime can be found at packagecontrol.io. Packages can also be found by hitting the CTRL + Shift + P keys (in Sublime) to open Package Control. Then, type install and enter the name of the package you are looking for in the box.

To enable modification of user settings, go to Preferences (top menu bar) > Settings and save the opening settings file either as ~./config/sublime-text/Packages/Default/Preferences.sublime-settings (recommended for first-time saving) or ~./config/sublime-text/Packages/User/Preferences.sublime-settings. Then, edit the desired settings: for instance, look for spell_check and set it to true to default-enable spell checking. Save the .sublime-settings file to apply changes.