Installation

Since Ontolearn is a Python library, you will need to have Python on your system. Python comes in various versions and with different, sometimes conflicting dependencies. Hence, most guides will recommend to set up a “virtual environment” to work in.

One such system for virtual python environments is python venv. Since the command comes together with python, you don’t need to install any external tool.

Installation via pip

Released versions of Ontolearn can be installed using pip, the Package Installer for Python. pip comes as part of Python.

pip install ontolearn

This will download and install the latest release version of Ontolearn and all its dependencies from https://pypi.org/project/ontolearn/.

Installation From Source

To download the Ontolearn source code, you will also need to have a copy of the Git version control system.

Install java and curl:

# for Unix systems (Linux and macOS)
sudo apt install openjdk-11-jdk
sudo apt install curl
# for Windows please check online for yourself :)

Once you have the done previous step, you can continue setting up a virtual environment and installing the dependencies. You may as well use your IDE interface to set up your venv.

->First download (clone) the source code

git clone https://github.com/dice-group/Ontolearn.git
cd Ontolearn

->Create a python virtual environment. (We are not using conda anymore)

python -m venv venv 
source venv/bin/activate # --> for Unix systems
# .\venv\Scripts\activate  --> for Windows

->Install the dependencies
```
pip install -r requirements.txt
```

Now you are ready to develop on Ontolearn or use the library!

Verify installation

To test if the installation was successful, you can try this command: It will only try to load the main library file into Python:

python -c "import ontolearn"

Tests

You can run the tests as follows but make sure you have installed the external files using the commands described here to successfully pass all the tests:

pytest

Note: Since Unix and Windows reference files differently, the test are set to work on Linux but in Widows the filepaths throughout test cases should be changed which is something that is not very convenient to do. If you really want to run the tests in Windows, you can make use of the replace all functionality to change them.

Download External Files

Some resources like pre-calculated embeddings or pre_trained_agents and datasets (ontologies) are not included in the repository directly. Use the command line command wget to download them from our data server.

NOTE: Before you run this commands in your terminal, make sure you are in the root directory of the project!

To download the datasets:

wget https://files.dice-research.org/projects/Ontolearn/KGs.zip -O ./KGs.zip

Then depending on your operating system, use the appropriate command to unzip the files:

# Windows
tar -xf KGs.zip

# or

# macOS and Linux
unzip KGs.zip

Finally, remove the .zip file:

rm KGs.zip

NCES data:

wget https://files.dice-research.org/projects/NCES/NCES_Ontolearn_Data/NCESData.zip -O ./NCESData.zip
unzip NCESData.zip
rm NCESData.zip

If you are getting any error check if the following flags can help:

unzip -o NCESData.zip
rm -f NCESData.zip

CLIP data:

wget https://files.dice-research.org/projects/Ontolearn/CLIP/CLIPData.zip
unzip CLIPData.zip
rm CLIPData.zip 

Building (sdist and bdist_wheel)

In order to create a distribution of the Ontolearn source code, typically when creating a new release, it is necessary to use the build tool. It can be invoked with:

python -m build

from the main source code folder. Packages created by build can then be uploaded as releases to the Python Package Index (PyPI) using twine.

Building the docs

The documentation can be built with

sphinx-build -M html docs/ docs/_build/

It is also possible to create a PDF manual, but that requires LaTeX to be installed:

sphinx-build -M latex docs/ docs/_build/

Simple Linting

Using the following command will run the linting tool flake8 on the source code.

flake8

Additionally, you can specify the path where you want to flake8 to run.

In the next guide, we explore about ontologies in Ontolearn and how you can modify them using axioms.