Local Machine Installation

Note: These installation instructions assume that you are familiar Python and using command line interfaces. If you would like to set this package up from scratch (no assumed experience), check out the more robust instructions here.

Requirements:

• OS: Linux or Mac/OSx
• Python 3.5 or higher (either system-level, or in an IDE)
• 400 MB free disk space
• Permission to perform sudo-level commands

Installation:

Installing epiphyte involves three broad steps: 1) installing and setting up a DataJoint docker container on the system-level, 2) download and launch the DataJoint MySQL server via docker-compose, and 3), installing the code package either on the system-level, or within a virtual environment (recommended). More information about the packages and software used by epiphyte can be found in the wiki.

1. Install and set-up the DataJoint docker container.

• Install the correct Docker for your OS.
• Verify the Docker installation:

docker --version
docker run hello-world

• Install Docker Compose.
• Verify the Docker Compose installation:

docker-compose --version

2. Download and launch the DataJoint MySQL server via docker-compose.

(Following is taken from DataJoint documentation.)

• Create a directory to store the docker-compose YAML file:

mkdir mysql-docker
cd mysql-docker
wget https://raw.githubusercontent.com/datajoint/mysql-docker/master/docker-compose.yaml
docker-compose up -d

• Download a package that enables access to your locally running MySQL server:

Linux:

sudo apt-get install mysql-client

Mac (via Homebrew):

brew install mysql@5.7
brew tap homebrew/services
brew services start mysql@5.7
brew services list
brew link --force mysql@5.7
mysql -V # verify the installation

• Test the server access. If there are issues, refer here.

mysql -h 127.0.0.1 -u root -p
Enter password: [type in your password here: default is "simple"]

3. Install Epiphyte:

There are two ways to install epiphyte -- via pip or cloning this repo.

Note: certain features of the DataJoint python package have not been updated for more recent releases of Python. Therefore, DataJoint and epiphyte require an older version of Python (<=3.9). For this reason, we recommend using a conda environment to install all needed packages and to develop from.

Install via git

• Clone the repository:

git clone git@github.com:mackelab/epiphyte.git
cd epiphyte

• Create a new conda environment suitable for the epiphyte requirements:

conda create --name epiphyte python=3.9.18 ipython

• Activate the conda environment:

conda activate epiphyte

• Install the needed dependencies using setup.py:

pip install .

• Verify the installation:

conda list

Or, install via pip in a conda environment

• Create a new conda environment suitable for the epiphyte requirements:

conda create --name epiphyte python=3.9.18 ipython

• Activate the conda environment and install epiphyte:

conda activate epiphyte
pip install epiphyte