API documentation

1 Deployment architecture

The general deployment architecture used for the MizoPol package is shown in Figure Figure 1 below.

Figure 1: Deployment architecture for the MizoPol package.

More precisely, a local continuously enriched features package is developed locally and periodically deployed on cloud via a docker-based images. The continuous deployment is done through a private Git-hub repository. This enables to deploy a set of endpoints that corresponds to the package exposed methods that are described in the following sections.

2 The mizopol interface module | installation

In order to use these endpoints while making total abstraction of the technicalities of the deployment, a module is developped and made freely downloadable from a public repository so that the endpoints be accessible from the user side.

The interface module will shortly be downloadble or cloned from the following public repository (I’m working on it):

https://github.com/mazenalamir/mizopol_api.git

Once downloaded, the folder containing the interface module shows the following architecture

.
├── README.md
├── mizopol
│   ├── __init__.py
│   ├── __pycache__
│   ├── g2sys_api.py
│   ├── plars_api.py
│   ├── pwpol_api.py
│   ├── rlars_api.py
│   └── utils_api.py
├── pyproject.toml
├── test_g2sys.py
├── test_plars.py
├── test_plars_doc.py
├── test_pwpol.py
├── test_rlars.py
└── uv.lock

The components of this folder are the following:

1. The MizoPol interface modules folder
This folder contains the modules that are designed to call the cloud deployed endpoints as shown in Figure Figure 1. For instance, in order to fit a plars model, the user would use the following import statement: 
from mizopol.plars_api import fit
2. A set of test python files
These files are provided as example of use of the available method in the MizoPol api. They cover the use of plars, g2sys (scripting version not the streamlit one), pwpol and rlars. Notice however that these files do not necessarily explore all possible parameters setting. To this end, reading the API documentation provided in the next section is necessary to fully exploit the capabilities of the MizoPol package.
3. The virtual environment uv utilities
This includes the uv.lock and the project.toml files. These files enables an easy creation of the virtual environment through the command: 
uv sync

It is possible that some of the test files make reference to dataset downloading. In such cases, the corresponding downloading instruction should be replaced by the user’s own files.

In the following section, the information needed to use the modules contained in the mizpol folder of the above shown architecture are detailed.