AramisAPI

A Julia API for the Aramis project.

Installation

The machine learning algorithms running in the background require Python version 3.6 with the pandas and scikit-learn packages, called from Julia through the PyCall package.

Since Python 3.6 is presumably not the default version on your machine, you have to tell PyCall where to find it. In a Julia REPL, run

import Pkg
Pkg.add("PyCall")

ENV["PYTHON"] = "<path to your python3.6 executable>"
Pkg.build("PyCall")

You can verify which version of Python is being used by PyCall with the commands

using PyCall
sys = pyimport("sys")
sys.version

Alternatively, you can set up a virtual environment using Anaconda (Miniconda installation link). To do so, download the script setup.jl and run it in the Julia REPL with

include("setup.jl")

or directly in the command line as

julia setup.jl

If successful, the installation script should output something like Using Python 3.6.15 ....

Once this is done, the Julia package AramisAPI.jl can be installed with

using Pkg
Pkg.add(url="https://github.com/GeeeHesso/AramisAPI.jl")

The installation can be tested with

using Pkg
Pkg.test("AramisAPI")

Usage

Load the package and start the API with

using AramisAPI
start_server()

The default IP is 127.0.0.1 and the default port 8080. After launching the server, the Swagger documentation page is available at http://127.0.0.1:8080/docs.

Docker container

The Docker file provided with the package can be used to build a container. Download the file and run

docker build -t aramis-api <path-to-the-docker-file>

The image created in this way is about 5 GB. To run the container, use

docker run -p 8080:8080 --restart=unless-stopped aramis-api

The API is then available at http://0.0.0.0:8080/docs.