flask_serving_eae

Simple API to serve dataset info and a ML model based on the Scikit-learn iris dataset

Goal

Here you can define business goals

Tested in Python 3.10 (should work with Python 3.6+, if not, open an issue)
- Students can add their own versions in which it was tested, create a Pull Request for this README.md)
(Optional) Create a virtualenv if you don't want to rewrite your environment with new library versions instead of the ones from Anaconda
Virtualenv:
- python -m venv .venv (check if the dir exists)
- Non-Windows: source .venv/bin/activate (you're going to see (.venv) in your prompt
- Windows: source .venv/Scripts/activate
Use make install or directly pip install -r requirements.txt

It's considered good practice to include running instructions to help other reproduce (run, test and evolve) your software.

Run the Flask Server using the following alternatives:

As a Python module: python iris_api.py (check the end of the file: the __main__ part)
Flask local server oneliner: export FLASK_APP=iris_api && export FLASK_DEBUG=1 && flask run --port=9000
Here you're defining environment variables and arguments like port to flask binary to know how to execute and which module iris_api to run
Finally, you can run it as a script in sh run_server.sh
(If you see permission denied errors) Give executable permissions to the shell script using chmod +x run_server.sh
Then, run sh run_server.sh

Run these examples in your browser to test main features of your Web Server

http://<your_host>:<your_port>/api/iris/classify?sl=5.1&sw=3.5&pl=1.4&pw=0.2
http://localhost:9000/api/iris/classify?sl=5.1&sw=3.5&pl=1.4&pw=0.2
Create executable permissions to the shell script chmod +x run_server.sh
Run with the following alternatives:
- As Python module python iris_api.py
- Flask local server oneliner export FLASK_APP=iris_api && export FLASK_DEBUG=1 && flask run --port=9000
- sh run_server.sh