3.0 The Data Set - WayneWang86/UDistrict-Real-Time-Fire-Response GitHub Wiki

Seattle Real Time Fire 911 Calls -- This is a real-time database record Seattle Fire Department’s 911 dispatches. It was created by the City of Seattle and can be found on the Seattle government data set website. An “observation” presents a single 911 fire call and the corresponding 911 response. Each “observation” shows the address, occurrence DateTime, incident type, specific location (latitude and longitude) as well as the incident record number of an incident. The strengths of the data set are it considers key values at stakes, such as location and incident type. These key values help users better interpret and organize information.

Besides, this data set can be used widespread for different stakeholders. For example, it can be used to project future fire responses and can be used by journalists to do media reports and by governments to do political reports. The weaknesses of the data set are that it is not diverse enough in terms of geographies. This data set only focuses on the Seattle area. Additionally, it can contain more desired values, such as the number of people. The duration of this dataset is from 2003 to now. For our project, we will only analyze the recent three year’s data (2017 - 2019). This dataset includes data of incidents’ addresses, types, DateTime, and locations. This dataset can be used to analyze the fire call frequency from time to time as well as identify the area with more frequent fire calls.

To clean and organize the raw data, we first extracted the most updated data through API and make it a data frame. Then we clean and reorganize the data frame within the two visualization files. For making aggregated data frames for both visualizations, we clean the data frame by eliminating all observations with NA values, then we use filter() to get the observations from 2017 to present. We also create two columns to contain the information of each observation’s year and quarter. For making heatmap, we will also filter the data frame by longitude, latitude and season(quarter) of users’ interest. For making the bar chart, we will group the data frame by quarter and filter the data frame by type of incidents based on users’ interests.

Attributes: 7 , Observations: 1.41M
Seattle Real Time Fire 911 Calls

Details about Data Set Variables: Data Dictionary