README.md

Collection of analysis performed using Python 🐍

Author

• Sreejith Sreekumar

The data for the analysis is about 650 MBs and has been put in the google live location shared here

https://drive.google.com/drive/folders/1Z7ZZFldy-XB1koU3JPW7oTQqs-jPGPMD?usp=sharing

Folder Organization

• htmls : Since the map outputs do not render on github, html output documents of previous successful runs
  have been placed here as a demo.
• scrap-code : Intermediate scrap code used in analysis - Clusterings which didn't give fruitful results / Time
• address_list_x : These files are address from geo-tags in json format. Since most service providers restrict decoding geo to address api limits to ~ 2000 calls per day, the decoded addresses have been cached and kept in these files.

Dependencies

1. pandas==0.21.0
2. matplotlib==2.0.2
3. folium==0.5.0
4. json==1.35
5. scikit-learn==0.19.1

• A requirements.txt file has been added in the folder. Please use pip install -r requirements.txt for installing the requirements

Important Note

• The data has to be downloaded and the variable data_folder in the notebooks has to be set to the downloaded path for the code to run correctly.

Notebooks

• Extended EDA for London
• Cluster Analysis for patterns in traffic
• HeatMap of Accidents which has been used for traffic correlation checks