Skip to content

Latest commit

 

History

History
63 lines (42 loc) · 1.5 KB

README.md

File metadata and controls

63 lines (42 loc) · 1.5 KB

TensorFlow JS KNN Dashboard

This small project aim to create a demo for KNN implementation in JS. There's a dashboard where you can upload your csv and then you can choose your features and label.

Install locally

This project is made with NextJS (React) and Redux plus Material-UI.

To start playing around clone the project and then run inside the folder:

npm run i

npm run dev


Code Style guide

Javascript

ES6 First

  • async/await
  • const & let
  • arrow function
  • map, filter, reduce, forEach
  • ; at the end

CSS

  • css inside every component/element named ComponentName.module.css
  • when imported import ComponentNameCss from './ComponentName.module.css'
  • class names inside the module could be whatever cause with next they will be renamed uniquely at build time

Component Rules

  • All elements and components should be in the functional way (no class) if possible
  • An element is a react component that doesn't contain another component
  • Inside every element/component we're going to have this header sections:
// GENERIC IMPORTS
import React from 'react';
const tf = require('@tensorflow/tfjs');

// CONTROLLERS
import { knn, error } from '../../controllers/TensorFlowManager';

// COMPONENTS & ELEMENTS
import ResetButton from '../../elements/ResetButton'

// STYLES
import MenuCss from './Menu.module.css'

// ACTIONS
import { removeTensors } from '../../redux/actions/tensorsActions'
...

// CONSTANTS & VAR
const k = 30;
  • Containers (?) needs to be evaluated