Installation

Software Requirements

• Python (we currently recommend Python 3.8)

• Git

Ackrep Setup:

1. (Optional:) Create new environment (e.g. using anaconda).

2. Create a new empty root directory for the project (e.g. “ackrep”)

   cd <root directory>

3. Clone ackrep_data

   git clone https://github.com/ackrep-org/ackrep_data.git

4. Clone ackrep_core

   git clone https://github.com/ackrep-org/ackrep_core.git

   Your project directory should look like this:

    <root_dir>/
    ├── ackrep/
    │   ├── ackrep_data/                   ← separate repository for ackrep_data
    │   │   ├── .git/
    │   │   └── ...
    │   │
    │   └── ackrep_core/                   ← separate repository for ackrep_core
    │       ├── .git/
    │       └── ...
    │
    └── erk/                               ← if this is your first visit here, dont worry about this.
        └── erk-data                         if you want to use pyerk functionality, checkout the links below
            └─ ocse/
                ├── .git/
                └── ...
   

   (PyERK documentation, PYERK repository)

5. Enter the ackrep_core-repo directory, run the following commands from there.

   cd ackrep_core

6. Install dependencies

   pip install -r requirements.txt

7. Setup ackrep

   pip install -e .

8. Validate Installation

   ackrep --version
   # Version 0.6.0
   

9. Setup the database

   python manage.py makemigrations
python manage.py migrate --noinput --run-syncdb

---

Setup advanced control theory algorithms

With Docker

Note

Installing docker comes with some challenges, especially for windows users. But the advantage to be gained is the ability to execute any piece of ackrep code in a deliberately set up environment with all necessary dependancies already installed. One can be fairely certain, that the code inside the environment will behave as intended even after a long period of time.

• Install Docker Engine

• Linux Users: consider these post installation steps

• Validate the installation with

  docker run hello-world

• If not already done, complete the above described ackrep setup. You do not need to install runner dependancies!

Now you are ready to execute all examples on your own machine:
Continue with running the code from the command line.

---

Without Docker

Warning

If you do not already have docker set up on your machine, this is a relatively fast way to get startet. Most of the examples should be possible to execute without the need for docker. But some of the Ackrep examples require a speacial environment to run in, therefore usually docker containers are used.

• Install Dependencies

  pip install -r requirements_runner.txt

  If the installation of slycot fails, refer to https://github.com/python-control/Slycot#installing or remove the packages slycot and casadi from the requirements file.

Now you are ready to execute some examples on your own machine:
Continue with running the code from the command line.