We provide two ways to use pyAMNESIA.
- If you want a lightweight tool that runs in a command-line interface, you may want to install pyAMNESIA as a standalone tool;
- If you want to use CICADA to run the tool and have a pretty GUI to guide you, then you may want to install pyAMNESIA as a CICADA module.
We advise that you use conda to create a virtual environment for the tool.
conda create -n pyamnesia python==3.8 # or python==3.7 conda activate pyamnesia
You only need to
git clone pyAMNESIA:
# clone the repo git clone https://gitlab.com/cossartlab/pyamnesia.git cd pyamnesia # install the dependencies sh requirements.sh
On Windows, you may not be able to use the
sh command. To solve this, you may either download Cygwin or write the commands in
pip install -r requirements.txt conda install -c conda-forge skan conda install -c conda-forge hdbscan
The installation relies on conda to install skan, which is the package we use to skeletonize images. If you cannot use conda, please install skan and hdbscan with
pip; we cannot guarantee that it will run smoothly though.
If you want to use pyAMNESIA as a standalone tool, no need for you to read the next section and you can read the how-to page to run the tool.
# using pip pip install -U pycicada # from source git clone https://gitlab.com/cossartlab/cicada.git
As for now, CICADA still is a work in progress, we advise that you clone the stable version we worked on and pip install some additional packages to make it run:
# install stable version git clone -b stable-pyamnesia https://gitlab.com/theodumont/cicada.git cd cicada # pip install some packages pip install -r requirements.txt pip install neo elephant
CICADA and pyAMNESIA are now installed and you can go to the how-to page to run the tool.