Bike_Sharing_Analysis

Analysis of bike-sharing usage patterns and trends over a year.

Overview

This bike-sharing analysis uses skills and knowledge learned from DataCamp, the "Python for Data Analysis" book, and "The Data Analysis Workshop" book. By closely examining a real-world dataset, the project aims to answer important questions and explore different aspects of bike sharing, such as how time and weather affect bike rides. It also tests hypotheses based on the visualizations.

Dataset

title: Bike Sharing Source: UC Irvine Machine Learning Repository

Link: https://archive.ics.uci.edu/dataset/275/bike+sharing+dataset

Description: Bike-sharing systems are a modern version of traditional bike rentals, with automated processes for membership, rental, and return. Users can rent a bike from one location and return it at another. There are over 500 bike-sharing programs worldwide, with more than 500 thousands bicycles. These systems are popular due to their positive impact on traffic, the environment, and health.

Project Goals

• Explore the data: Checked the data, preprocessed it, and transformed it into a human-readable format.

• Analyzing seasonal impact on rides: it looks at how different seasons affect bike use. It helps us see if bike rides go up or down in different weather conditions, like summer or winter. This information can be used to improve bike-sharing services based on seasonal changes.

• Hypothesis testing: Estimating the average number of registered bike rides involves checking if the observed ride count matches a predicted average. By comparing actual data to expectations, we can see if any differences are statistically significant, which helps assess if bike-sharing programs meet their goals or if there are important variations.

• Analysis of weather-related features: Examining the correlation between continuous data such as temperature, season, humidity, and bike usage (both registered and casual) involves analyzing how these variables relate to each other.

• Time series analysis: checking if data is stationary and using methods like rolling statistics, last values, and decomposition to address non-stationarity. These techniques help in understanding and forecasting trends and patterns over time.

  Project Output

  • Pyhon scripts for each step of project:

    Script Sequence

    1. Data Directory:
       • 01_Processing_data.py
    2. visualization Directory:
       • 01_Preprocess_temporal_and_weather_feature.py
       • 02_Hypothises testing.py
       • 03_Analysis_of_Weather_Related_Features.py
       • 04_Time series analysis.py

  • Visualization Figures

  • Explanation of each section of code as comments

This project demonstrates how data analytics skills can be used to find useful insights in bike-sharing. It shows the importance of making decisions based on data and highlights expertise in data analysis.