UDAVA

Unsupervised learning for DAta VAlidation.

Installation and setup

Choose option A or B to start the Udava service on your computer.

A) Running Udava in Docker

docker build -t udava -f Dockerfile .
docker run -p 5000:5000 -it -v $(pwd)/assets:/usr/Udava/assets -v $(pwd)/.dvc:/usr/Udava/.dvc udava

B) Run Udava directly on host

You can install the required modules by creating a virtual environment and install the requirements.txt-file (run these commands from the main folder):

mkdir venv
python3 -m venv venv
source venv/bin/activate
pip3 install -r requirements.txt

Start the server by running:

python3 src/api.py

Usage

GUI

The GUI is available at http://localhost:5000/. Follow the instructions in the GUI to create models and upload data for inference.

API

The API is available at http://localhost:5000/.

POST /infer

JSON

The input should look like this:

{
  "param": {
    "modeluid": "3a7ee233-2380-4420-9e1d-246932bdede4"
  },
  "scalar": {
    "headers": ["time", "memory_used"],
    "data": [
      [1718725511, 1201184768],
      [1718725514, 1201840128]
    ]
  }
}

Explanation:

param.modeluid: The unique identifier (UUID) of the model used.
scalar.headers: An array of strings representing the data columns (in this case, time and memory_used).
scalar.data: A 2D array with each inner array representing a data point:
- The first value is a Unix timestamp.
- The second value is the memory used in bytes.

The JSON can be sent to the API using curl:

curl -X POST -H "Content-Type: application/json" -d @data.json http://locahost:5000/infer

CSV

CSV data can be sent to the API using curl:

curl http://localhost:5000/infer -F file=@data.csv -F model_id=151d2394-7654-4958-9e82-174c7198368c

Make sure that the CSV contains the same columns as the model expects.

SINTEF-9012/Udava