CS 560: Midterm Team Enigma

Dataset

Information about our Dataset

Fire Department Calls for Service

Data comes from here, from DataSF.

License: Open Data Commons Public Domain Dedication and License

Accessed on March 1, 2020.

Data

There are a total of 34 columns and about 5,208,517 rows in this dataset.

Possible Columns of Interest

  1. Neighborhoods - Analysis Boundaries: Which San Francisco neighborhoods the services calls were from. Snippet of possible values:
    • Bayview Hunters Point
    • Bernal Heights
    • Chinatown
    • Excelsior
    • Golden Gate Park
    • Inner Sunset
    • Japantown
    • Lone Mountain/USF
    • Noe Valley
    • Tenderloin

  2. Call Type Group: The call's group type. Possible values:
    • Alarm
    • Fire
    • Non Life-Threatening
    • Potentially Life-Threatening

  3. Call Type Category: More specific information about the call type. Possible values:
    • Confined Space / Structure Collapse
    • Electrical Hazard
    • Elevator / Escalator Rescue
    • Medical Incident
    • Suspicious Package
    • Traffic Collision
    • Train / Rail Fire
    • Train / Rail Incident
    • Transfer
    • Vehicle Fire
    • Water Rescue

  4. Received DtTm: The data and time of when the call was received. Format:
  5. YYYY Month DD HH:MM:SS AM/PM

Wrangling

Trends of Number of Called Deemed "Life Threatening": Data was wrangled first prior to download, through the editor provided on the data provider's website. There, we were able to filter to download only the calls that were part of the call group "Life-Threatening", and only download data between the years of 2013 and 2019. That data was then wrangled in Excel, where we removed unecessary columns. Finally, additional filtering, namely excluding all but the top 15 neighborhoods in terms of number of life-threatening calls, was done in Tableau.