optimizer

Template for an optimizer to be deployed into production. A simple knapsack problem is built as a sample.

Set Up

Create a Python virtual environment. Install the virtualenv library.

pip install virtualenv

Create the venv.

virtualenv venv

Activate the venv.

source venv/bin/activate

Install requirements.

pip install -r requirements.txt

Deploy

Locally

First, set your PYTHONPATH.

export PYTHONPATH=.

Run the app.

uvicorn main:optimizer --host localhost --port 8080

or

python3 main.py

Make requests to http://localhost:8080.

Docker

Build the image with the Dockerfile recipe and name it optimizer.

docker build -t optimizer .

Run the optimizer image in the background (-d), name it optimizer-app and expose port 8080.

docker run -d --name optimizer-app -p 8080:8080 optimizer

Make requests to http://localhost:8080.

Use

Make the following request:

curl -L -X POST 'http://localhost:8080/knapsack/' \
-H 'Content-Type: application/json' \
--data-raw '{
    "items": [
        {
            "name": "Laptop",
            "weight": 4,
            "value": 3
        },
        {
            "name": "Cellphone",
            "weight": 3,
            "value": 5
        },
        {
            "name": "Headphones",
            "weight": 2,
            "value": 3
        },
        {
            "name": "Charger",
            "weight": 4,
            "value": 1
        },
        {
            "name": "Extension",
            "weight": 5,
            "value": 4
        },
        {
            "name": "Camera",
            "weight": 5,
            "value": 5
        },
        {
            "name": "Tablet",
            "weight": 6,
            "value": 5
        },
        {
            "name": "Case",
            "weight": 2,
            "value": 1
        },
        {
            "name": "Glasses",
            "weight": 4,
            "value": 1
        },
        {
            "name": "Earphones",
            "weight": 2,
            "value": 2
        },
        {
            "name": "Ultrabook",
            "weight": 5,
            "value": 7
        },
        {
            "name": "Sunglasses",
            "weight": 2,
            "value": 3
        },
        {
            "name": "Magazine",
            "weight": 3,
            "value": 1
        },
        {
            "name": "Book",
            "weight": 2,
            "value": 4
        },
        {
            "name": "Powerbank",
            "weight": 2,
            "value": 2
        },
        {
            "name": "Sunscreen",
            "weight": 2,
            "value": 5
        }
    ],
    "capacity": 20
}'

You should get the following response:

{
  "selected_items": [
    "Cellphone",
    "Headphones",
    "Earphones",
    "Ultrabook",
    "Sunglasses",
    "Book",
    "Powerbank",
    "Sunscreen"
  ],
  "total_value": 31.0,
  "total_weight": 20.0
}