UBER-EATS RESTAURANTS BIG DATA MENU PRICE ANALYSIS

Project Overview

Big data analytics for delivery companies is always very necessary to uncover customer trends and patterns. Trends help these companies in optimizing their services to ensure a smooth flow of operations between riders as well as forecast future demand of various products and services.

This project aimed to analyze the menus of various restaurants that use uber-eats courier as their supply company. As this is big data, the use of OLAP-DATABASES comes in handy to perfrom heavy analytics using SQL. These databases offer:

• Scalability: Can easily manage massive amounts of data efficiently.
• Low latency: Optimized for complex analytical queries due to their latency structure and optimized data structures and indexing.
• Batch processing: Good for large-scale data processing tasks efficiently.

Uber-Eats customers also have the right to investigate the prices of various items on the menu to find the cheapest and most expensive restaurants to save on expenditure. The data was sourced from kaggle and can be accessed by using the following link

Objectives

1. Investigate the number of restaurants that use UBER-EATS.
2. Investigate the top restaurants with the largest number of items.
3. Investigate the price range of items for the restaurant with the highest number of items.
4. Investigate the average price of items across various restaurants
5. Investigate the most expensive restaurant.
6. Snacks items analysis. Analyze the price of snack items across various restaurants.

Findings

Future steps

1. Investigate the correlation between location i.e latitude & longitude and menu-prices.
2. Outsource data for customer orders and analyze consumer purchasing trends.

Getting started

All the requirements to run this project are listed in the file requirements-file together with the library versions.

1. Download the zip-file from kaggle
2. Make a project directory using the mkdir command on the terminal and extract the filesin the zip file.
3. The files are in zip formart having a size of 896 mb.
4. Follow along through the parquet convert file which uses apache pyarrow.
5. Load the parquet file into an OLAP database namely DuckDB in this case for quick analytical queries.