Atlas7 / visualize-traffic-accidents-in-uk

Data Science, Visualization, and Predictive Analytics: Kaggle Dataset - 1.6M accidents & traffic flow over 16 years

Visualize Traffic Accidents in UK

View the following notebooks on nbviewer:

• Visualize Traffic Accidents in UK with Datashader
• other notebooks

All notebooks are stored at the notebooks directory.

---

This repository contains a series of data science notebooks that performs data science, visualization, and predictive analytics against the kaggle Dataset: 1.6M accidents & traffic flow over 16 years - Visualise and analyse rich data on traffic demographics

All notebooks are stored under the notebooks directory.

---

Table of Content

• Setup Instructions
• Notebooks

---

## Setup Instructions

Download Data and Setup Input Directory

Download the zip file 1-6m-accidents-traffic-flow-over-16-years.zip from the kaggle website, and extract it to somewhere suitable on your local machine. I assume it goes to your ~/Downloads

At the root of the repository, create a directory input and download (and extract) all the flat files there.

Step 1: Create input directory:

$ mkdir input

Step 2: Move the 1-6m-accidents-traffic-flow-over-16-years.zip to thie input directory

$ mv ~/Downloads/1-6m-accidents-traffic-flow-over-16-years.zip ./input

Step 3: extract all files within the same input directory:

$ cd input
$ tar -xvzf 1-6m-accidents-traffic-flow-over-16-years.zip`

Once you have the zip file extracted you should see some .csv files.

Local_Authority_Districts_Dec_2016.geojson
accidents_2005_to_2007.csv
accidents_2009_to_2011.csv
accidents_2012_to_2014.csv
accident_coords_update.ipynb
ukTrafficAADF.csv

Setup conda environment

$ conda env create -f environment.yml

Activate conda environment

$ source activate
$ source activate ds2

Start Jupyter Notebook

At the repository root directory:

$ jupyter notebook

Navigate to the notebooks directory, and pick the notebook you'd like.

---

Notebooks

All notebooks are stored under the notebooks directory. The Notebook README contains links to view the notebooks via the Jupyter nbviewer tool.