Graphs have many applications in computer science and software engineering, because they are a general model of relatedness. This repository is a collection of examples where something in software engineering has a useful and interesting graph structure that is straightforward—and fun—to visualize programmatically. Graphs are extremely versatile, so these examples are necessarily incomplete in the extreme.
These graphs should not be confused with the other thing called a graph, nor with plots (which relate to that other thing).
Currently all examples are in Python, with all the code in Jupyter notebooks.
builtin_types.ipynb- inheritance hierarchy of Python builtinsexceptions.ipynb- inheritance hierarchy of Python exceptionsfibonacci.ipynb- relationship of Fibonacci subproblemsgit_graph.ipynb- source control commit graphsnested_sequences.ipynb- heterogeneous nested collectionsnested_tuples.ipynb- homogeneous nested collectionsyaml.ipynb- structured data
Clone the repository and go into the cloned directory:
git clone https://github.com/EliahKagan/codegraph
cd codegraphThen you can install dependencies using conda or using pipenv.
To use conda, make
sure conda is installed (if not, I suggest
Miniforge). Then run:
conda env createTo use pipenv, make sure pipenv is
installed.
Also install Graphviz on your system (conda takes
care of that but pipenv doesn't). Graphviz 3.0.0 or higher is recommended, to
avoid scaling problems.
(Using conda instead of pipenv may be the easiest way to get a new version
of Graphviz on some systems.)
Then run:
pipenv install -dIf you already have an application for running code in Jupyter notebooks that
you want to use, such as VS Code with the
Jupyter
extension,
then you can omit the -d.
As for installing Graphviz, most systems' package managers will install it, and
the package name is usually Graphviz. If you can't find it, look into what
package provides the dot command. On Windows, if you want to install Graphviz
via a package manager, you can use scoop.
First, go to the directory where you ran conda or pipenv before and
activate the environment you created.
If you used conda, run:
conda activate codegraphIf you used pipenv, run:
pipenv shellThen run JupyterLab:
jupyter labIf you already have another application you prefer for opening and running code in Jupyter notebooks, you can use that.
In particular, you can use VS Code with the Jupyter Notebooks
extension.
If you do, make sure the environment you created (with conda or pipenv) is
selected. If VS Code offers to install ipykernel or any other dependency, you
should decline and select the appropriate environment.
An old, simpler version of some parts of this project is available on the simple branch, and may be of some interest.
Some parts of this project were tested by, and were improved based on input from, David Vassallo. Furthermore, significant parts were inspired by code I wrote for our project palgoviz, and some code to create nested tuple structures is derived from code there.