Developing Data Products - Project for Coursera Data Science Specialication - Auto Theft in Toronto

Parag Sengupta
April 24, 2018

Introduction - Auto Theft in Toronto

This presentation is for a project on Developing Data Products as part of the Coursera Data Science Specialization.

Analysis on: Auto Theft in Toronto (2014-2017).
13,480 data records of which 13,435 pertains to the 2014-2017 period.

Data Source: Toronto Police Service, Public Safety Data Portal:
http://data.torontopolice.on.ca/datasets/auto-theft-2014-to-2017

Related links and file locations for this presentation
App on shinyapps.io:
https://paragsengupta.shinyapps.io/gtaautotheft/

Source codes (Shiny ui.R and server.R) on GitHub repository:
https://github.com/paragsengupta/DevelopingDataProductsProject

Presentation & Functionality on GitHub repository:
https://github.com/paragsengupta/DevelopingDataProductsProject/myPresentationSlidy.Rmd
(R Markdown Slidy Presentation was used instead of R Presentation to create this pitch as RStudio v1.1.442 gave a message stating appshot of Shiny app objects - a key feature in this presentation - is not yet supported. As a continuation, a web version using R Presentation was created with Shiny app snapshort and published with RPubs)
RPubs: http://rpubs.com/paragsengupta/AutoTheftGTA

Shiny App: Functionalities

Main Panel:

  • Main Menu:
    • Map
    • Chart
    • Data Table
  • Main Display Zone: below the Main Menu

SideBar Panel:

  • Slider to choose the year range for the Interactive Map
  • Dropdown to choose whether to see All Neighbourhoods or Pick One
    • If Pick One chosen, Dropdown to choose the specific Neighbourhood by name
  • Radio buttons for choose the Chart type to display in Main Display Zone

Slidy App Screenshot

Since R Presentation is not compatible with real time Shiny app run in slides, a snapshot of the actual app in running is shown here.
plot of chunk unnamed-chunk-1
Code: shinyAppDir(“D:/Professional_n_Knowledge/Data Science John Hopkins/9 Developing Data Products/CourseProject/GTAAutoTheft/”, options = list(height=550))

To see the Shiny app actually running inside the presentation, please view “myPresentationSlidy.Rmd” from https://github.com/paragsengupta/DevelopingDataProductsProject/

Key Data Points (Variables) in the Auto Theft Data

  • Neighbourhoods:

    • 140 different neighbourhoods included in the data
      • West Humber-Clairville has the highest auto theft occurrences
  • Premise Types:

    • 5 premise types: 3 indoor, 1 outdoor and 1 for everything else
    • Apartment
    • Commercial
    • House
    • Outside
      • Highest auto theft occurences reported from Outside
    • Other
  • Month and Day of Occurrence:

    • Friday reported the highest auto theft in the week
    • October is the month reporting the highest auto thefts in the year