Skip to content

Latest commit

 

History

History
53 lines (31 loc) · 3.15 KB

README.md

File metadata and controls

53 lines (31 loc) · 3.15 KB

Bergen Bysykkel 2024: Python Edition - Cleaning, Predicting, and Mapping (Part 1)

This is Part 1 of a two-part analysis exploring Bergen's bicycle rental trends, now implemented in Python! In this installment, we perform data cleaning, build basic predictive models, and end with an interactive map showcasing bike traffic across stations. Part 2 will delve deeper into advanced methods, integrating weather data for precision modeling and actionable insights.

Data Sources

  • Bysykkel Data: Historical bike rental data provided by Bergen Bysykkel.
  • Weather Data: Hourly weather observations obtained from SeKlima.

Highlights of Part 1

  1. Python-Powered Data Cleaning:

    • Combined hourly bike rental data for 2024 into a clean and structured format.
    • Addressed missing values, ensured consistent timestamps, and added derived features (e.g., weekday, hour).

  2. Foundational Predictive Modeling:

    • Utilized Python's scikit-learn library for linear regression to predict hourly rentals by station, weekday, and time.

  3. Interactive Mapping:

    • Created an engaging map using Python's folium library to visualize station activity, complete with color-coded markers and popups.

What’s Coming in Part 2?

  • Integration of weather data to evaluate its influence on bike rentals.
  • Advanced predictive models leveraging machine learning techniques.
  • Comparative analysis of model performances.

Stay tuned as we transition from foundational techniques to advanced predictive analytics in Python!

Getting Started

Requirements

  • Python version >= 3.8
  • Required libraries: pandas, numpy, matplotlib, seaborn, scikit-learn, folium, etc.

Inspiration and Acknowledgments

The initial problem framing and directions for this project are inspired by the BAN400 course at the Norwegian School of Economics (NHH). To my knowledge, the earliest solution to this exam available on GitHub is:

Building on these foundations, this project transitions the analysis to Python, leveraging tools like pandas, scikit-learn, and folium to provide an enhanced and updated perspective using 2024 data.

A notable feature in this implementation is the use of Python's folium library for geospatial visualization. This approach mirrors the functionality of the leaflet package in R, leveraging free OpenStreetMap data to create an interactive mapping experience, complete with zoom controls, hover effects, and categorical markers.

Legal and Ethical Notes

  • This project references the BAN400 preparation material as a learning guide while ensuring originality in its implementation.
  • All work is aligned with NHH’s academic integrity guidelines and complies with the Bergen Bysykkel API's open-source licensing.