A Century of Airplane Crashes

Laine Gambeta
Geovisualization Project, @RyersonGeo, Fall 2019

Tableau is an exceptionally useful tool in visualizing data effectively.  It allows many variations of charts in which the software suggests the best type based on data content.  The following project uses a data-set obtained from the National Transportation and Safety Board identifying locations and details of plane crashes between 1908-2009. The following screenshot is a final product and a run through of how it was made.

Map Feature:

To create the map identifying accident location, a longitude and latitude is required.  Once inputted into the Columns and Rows, Tableau automatically recognizes the location data and creates a map. 

The Pages function is populated with the date of occurrence and filtered by month in order to create a time animation based on a monthly scale. When the Pages function is populated with a date the software automatically recognizes a time series animation and creates a time slide.

The size of the map icon indicates the total number of fatalities at a specific location and time.  To create this effect, the fatalities measure is inputted into the Size function.  This same measure is inserted into the label function to show the total number of occurrences with each icon appearance.

When you scroll over the icons on the map the details of each occurrence appear.  To create this tool, the measures you want to appear are inserted into the Details function.  In this function, Date, Sum Aboard, Sum Fatalities, Sum Survivors, and Summary of accident appears when you scroll over the icon on the map.

Vertical Bar Chart Feature:

To create the vertical bar chart you must insert the date on the Y axis (columns), and the X axis (rows) with people aboard and fatalities.

Next, we must create a calculation to pull the number of survivors by subtracting the two measures.  To do so, right click on a column title cell and click create calculated field.  Within this calculation you select the two columns

you want to subtract and it will populate the fields. We will use this to identify the number of survivors.

The next step is creating a dual- axis to show both values on the same chart.  Right click one of the measures in the rows field and click dual-axis.  This will combine the measures onto the same chart and overlap each other.

Following this we need to filter the data to move along the animation by month.  It tallies the monthly numbers and adds it to the chart. In order to combine the monthly tallies to show on an annual bar chart, the following filters are used.  First filter by year which tallies the monthly counts into a single column on the bar chart.  The Page’s filter identifies the time period increments used in the time slider animation, this value must be consistent across all charts in order to sync.  In this case, we are looking at statistics on a monthly basis.

To split the colours between green and red to identify survivors and fatalities, the Measure Names (which is created automatically by Tableau) is inserted into the colour function.  This will identify each variable as a different colour.

When you bring your mouse over top the bar chart it selects and identifies the statistics related to the specific year.  To create this feature, the measures must be added to the tooltip function and formatted as you please.

Horizontal Bar Chart Feature:

The second bar chart is similar to the previous one.  The sum of fatalities is put in Columns and the Date is put in Rows to switch the axis to have the date on the Y axis.  The Pages function uses the same time frame as other charts calculating monthly and adding the total to the bar chart as the time progresses.

Total Count Features:

To create the chart you must insert the date on the Y axis (columns), and the X axis (rows) with people aboard and fatalities.

Adding in running counts is a very simple calculation feature and is built into Tableau.  You build the table by putting the measure into the text function, this enable‚Äôs the value to show as text and not a chart.  You will notice below that the Pages function must be populated with a date measure on a monthly basis to be consistent with the other charts.   

In order to create the running total values, a calculation must be added to the measure.  Clicking the SUM measure opens the options and allows us to select Edit Table Calculation.  This opens a menu where you can select Running Total, Sum on a monthly basis.  We apply this to 3 separate counters to total occurrences, fatalities, and survivors.

Pie Chart Feature:

Creating a pie chart requires the following measures to be used.  Under the marks drop down you must select pie chart.  This automatically creates a function for angular measure values.  The fatality and survivor measures are used and filtered monthly.  The Measure Values which is automatically created by Tableau identifies the values of these measures and is inputted into the Angle function to calculate the pie chart.  Again, the Measures Names are inputted into the colour function to separate the values by fatalities and survivors. The Pages function is populated with date of occurrence by month to sync with the other charts.

Lastly, a dashboard is created which allows the placement of the features across a single page.  They can be arranged to be aesthetically pleasing and informative.  Formatting can be done in the dashboard page to manipulate the colors and fonts.

Limitations:

Tableau does not allow you to select your map projection. Tableau Online has a public server to publish dashboards to, however it does not support timeline animation. Therefore, the following link to my project is limited to selecting the date manually to observe the statistics.

https://prod-useast-a.online.tableau.com/t/lainegambeta/views/ACenturyofAirplaneCrashes/Dashboard2?:origin=card_share_link&:embed=n