Visualization API

Table of Contents

• Table of Contents
• Core Concepts
• Getting started
• API
• Continous Integration & Deployment
• Datasets
  • BIRAFFE2

Core Concepts

Visualization API purpose is to fetch datasets and prepare them for visualization using frontend React + D3.js app.

It is a koa app that is meant to be run as lambda fetching from S3 in production environment using serverless and serverless-http, however other than testing purposes it is not live due to financial reasons and should be used locally together with a simple JSON serving server like visualization-data-bucket.

It is also hosted in Heroku for convenience sake Heroku preview.

Getting started

Clone the repository and go inside it

git clone https://github.com/kuzmialke-and-friends/visualization-api
cd visualization-api

Check node&npm version and install dependencies

nvm use
npm i

Start development server

PORT=3000 DATA_BACKEND=https://knf-visualization-data-bucket.herokuapp.com/assets/biraffe npm run start:dev

Build and run production

npm run build
PORT=3000 DATA_BACKEND=https://knf-visualization-data-bucket.herokuapp.com/assets/biraffe npm start

Run test

npm test

Visualization API will be running on:

http://localhost:3000/

API

TODO: implement API specs

Endpoints:

Datasets

/datasets/:id

Query parameters:

• limit - default 2, limits number of subjects we sent data to

Note: does not work on Heroku due to memory constraints, TODO: fix memory limitation on limit function

Examples:

https://knf-visualization-api.herokuapp.com/datasets/ghost?limit=15
https://knf-visualization-api.herokuapp.com/datasets/jump

Static Maps

/static-map/:id

Examples:

https://knf-visualization-api.herokuapp.com/static-map/ghost

Visualization API will output data from dataset in format:

{
subjects: Subjects;
supportedVisualizations: VisualizationType[];
}

Health

/health

Returns 200 with Ok body response.

https://knf-visualization-api.herokuapp.com/health

Continous Integration & Deployment

Code is written and meant to be run as Lambda, however it is currently deployed to Heroku using their Free Dynos.

This app uses Travis for testing each change done to codebase automatically and as early as possible. Continuous Deployment follows the testing that happens during CI and pushes changes to production system.

Stages:

Test

• run tests
• run linting

Build

• builts app

Deploy

• currently disabled
• automatically deploys lambda to AWS

Datasets

BIRAFFE2

The "BIRAFFE2: The 2nd Study in Bio-Reactions and Faces for Emotion-based Personalization for AI Systems" dataset is part of the work carried out by the AfCAI Research Group in Kraków, Poland.

Source: BIRAFFE2