Smart Dispenser Rest Network API and data analysis

using Windows10, Python3.8, MySQL, Flask, SQL Alchemy, Marshmallow, Keras.

Prerequisites

Set up a mysql server and replace the line:
app.config['SQLALCHEMY_DATABASE_URI'] = 'mysql+pymysql://janesh:password@localhost/testdatabase'
with your own modified version corresponding to your database settings:
app.config['SQLALCHEMY_DATABASE_URI'] = 'mysql+pymysql://<username>:<password>@<server ip>/<database name>'

Quick Start

# install dependencies
$ pip3 install -r requirements.txt

# create database
$ create_db.py

# start server
$ server.py

# start client in another terminal (while server is running)
$ client.py

# execute data analysis (while server is running)
$ data_analysis.py

# run machine learning to predict days until dispenser fluid level is depleted
$ machine_learning.py

View Data

To view the data stored on the mysql databse you can run the following sql query after inserting your database name:

use <database name>;
select * from dispenser;
select * from dispenser_data;

Optional

Allow any device on LAN to connect to server API

You need to find the ip address of the machine the server is running on.
Open a terminal and enter the following command:

# for windows
$ ipconfig

# for linux
$ ifconfig

Find the IPv4 address for you wireless LAN and copy it, it will be formatted like this e.g. 192.168.0.52
Head to the client.py file and find the one line that contains localhost at the bottom of the file.
Replace localhost with the copied ip address. Also replace the line at the top of the file in data_analysis.py.

Any device currently connected to your local area network(LAN) via wifi will now be able to make requests to the server.

Simulate multiple dispensers

Copy paste the client folder multiple times and run each client.py in their own terminal window.
Ensure that server is currently running and that each new client folder copied has the config.json file set to it's default configuration:

# config.json
{
    "created": "",
    "device_id": null,
    "location": "",
    "name": ""
}