Table of Contents

• About the Project
• Getting Started
  • Prerequisites
  • Installation
• Usage
• toDo

About The Project

Solution for the mushroom-challenge

Getting Started

Get the code running.

Prerequisites

You need the following software

• nodejs & npm download
• mongoDB download
• angular

npm install angular -g

Installation

1. Clone the repo

git clone https://github.com/JanKru/mushroomClassifier.git

2. Navigate to mushroomClassifier/backend

cd ./mushroomClassifier/backend

3. Create a .env-file (like the example below)

NODE_ENV=development
SERVER_PORT=4041
MONGO_HOST=mongodb://localhost/mushroomClassifier

JWT_SECRET=SECRET#1234
JWT_EXP=20m

HIDDEN_LAYERS=[200]
ITERATIONS=20000
ERRORTHRESH=0.005
LEARNING_RATE=0.3
TIMEOUT=120000

THRESHOLDTOBEPOISONOUS=85

4. Install NPM packages

npm install

5. If you want to run backend: unit test

npm test

6. Navigate to the frontend diretory

cd ../mushroomClassifier/frontend

7. Install NPM packages

npm install

8. Start the frontend

ng serve --open

9. Start the backend (backend dir)

node index.js

10. Open Webpage and register a new user

Usage

Start the server (node index.js) and frontend (ng serve --open) Be sure mongoDB is running

When your work on Mac your linter may marks the linebreaks. Go to .eslintrc and change linebreak style to unix

 "linebreak-style": ["error", "unix"]

ToDo

• solve the responsive problem in modal
• Capsules the individual modules correctly
• Write more tests, especially for the neural network. Possibly train with a smaller amount of data first and then test it