Capstone Project

Predict the risk given a description of the task

Docker Installation

Pre-Requisite

• For windows, install Docker desktop

• For Ubuntu, Follow instructions here

Build Docker image

Go to the containing folder and run below command in command line

`docker build -t risk-predictor-app .`

_{Note: Add sudo if you are running Docker with root}

Run Docker Image

`docker run -p 5000:5000 -t risk-predictor-app`

Ubuntu OS

Pre-Requisite

• Python 3

• pip3

  _{install using sudo apt-get install python3-pip}

Installation

! Recommended to install virtual environment for python. (Might be pre-installed)

1. Virtual Environment [optional]

1. Installation

   sudo pip3 install virtualenv

2. Create virtual environment on repo

   virtualenv -p python3 ./risk_predictor_venv

3. Activate the virtual environment

   source risk_predictor_venv/bin/activate

2. Install dependencies

You can install dependencies from the requirements.txt file

`pip -r requirements.txt`

3.Install Postman [for testing]

Download Postman from their downloads page for your OS

_{Note: Postman collection is available in the repository for importing}

4. Server.py

In the last line, change

app.run(port=5000,host='0.0.0.0', debug=True)

to use the localhost address 127.0.0.1

app.run(port=5000,host='127.0.0.1', debug=True)

Run the Application Server

`python3 server.py`

APIs Available

When running docker : use host address: 172.17.0.2:5000

when running as standalone python, you can use 127.0.0.1:5000

E.g.: 127.0.0.1:5000\isalive or 172.17.0.2:5000\isalive

API	type	Description	Expected Request Body	Expected Response
/isalive	GET	Checks if the service is available.		{"status": "Live"}
/predict	POST	Description of task is passed as JSON and the service returns the accident level category from 0 to 4	{"text":"This is a sample description of task"}	{"prediction":"Don't worry, there is little risk (Category II)"}

Web Application

Access the application at address localhost:5000