DataViz Makeover 3

A geo-spatial and temporal visualization of armed conflict events in South East Asia.

Critiques and Suggestions

Original Visualization

#	Critique	Suggestion
	Clarity
1	The scale of the Y-axis of the line charts are not synchronized, which can lead to misinterpretation when comparing the number of events. For example, on first glance, it would seem like in 2020, there are more Explosions events compare to Battles but it is in fact the other way around	Synchronized all the Y-axis to have the same scale, or better yet, combine all 6 line charts into a single chart
2	Title of the chart on the right does not reflect the content of the chart, which is a count of events per event type	Rephrase the title of the line charts to provide better details
3	“Count of Sheet1” is meaningless in explaining the Y-axis and redundant	Remove “Count of Sheet1” from the visualization
	Aesthetic
1	The dots in the map do not have any borders and are not transparent, which makes it difficult to discern events which are overlapped or near to each other	Add borders for each point and make them slightly transparent
2	The line charts are not color coded and ordered in a different order compare to the provided legend, which can cause confusion	Color code the lines such that they follow the color of the Event Types in the map
	Interactivity
1	The visualization does not provided a filter for Event Date and Event Type	Add a filter for Event Date and Event Type
2	There are no interactions between the map and the charts on the right	Add interactions such as selecting a point in the map will highlight the line chart pertaining to the event type of the selected point

Proposed Design

• Because the locations in the dataset are not standardized, the events can only be grouped to the Country level, hence choropleth map is not suitable to represent the data. A better way to visualize the spatial data is by using both a heatmap, for overview of the location and density of the armed conflict events, and a point distribution map for more detailed view of individual events. Users will have the option to switch between the 2 types of map.
• Instead of having separate line charts similar to the original visualization, combining the line charts into a single chart will make it easier to do comparison across event types. The lines will be color coded in the same color as the map’s legend.
• Putting the legend at the top right corner of the map will allow the legends be referenced when using both the map and the line charts.
• The dataset contains event dates to the date level, hence some insights maybe lost when the events are grouped by year. Having a Date Granularity parameter will give users the option to group or drilldown by events dates, potentially providing additional insights.
• The maps and the line charts are highly correlated so a helpful interaction would be to highlight the same event type on both the map and the line chart when user clicks on either a certain events in the map or a line in the chart.

How-to Guide

Point Distribution Map

1. Open Tableau and add the data file as a data source
2. Go to Sheet1 and rename the sheet to Point
3. Add Longitude (generated) into the Columns card.
4. Add Latitude (generated) into the Rows card two times and make them Dual Axis.

5. For the first Marks card, set the mark type to Map and drag-and-drop Country into Detail. Set Color to #b4b4b4 with 40% opacity.

6. For the second Marks card, set the mark type to Circle and drag-and-drop:
   • Event Type to Color.
   • Latitude to Detail and set Measure to Average.
   • Longitude to Detail and set Measure to Average.
   • Event Id Cnty to Detail.

7. Set the colors for Event Type as follow, using the color picker tool provided by Tableau.

8. Set the border color for the Circle mark to #555555 and opacity to 60%.

9. Add Country, Event Type and Event Date (by Year) to the Filters card. Right-click on each one of them and select Apply to Worksheets > All Using This Data Source.

10. On the Toolbar, change the map background color by selecting Map > Background Maps > Dark. By now, the visualization should look like this:

11. Create a Parameter named Map Selection with the following properties:

12. Show the Map Selection parameter and set the value to Point Distribution.

13. Create a Calculated Attribute named Selected Map with the formula below:

[Map Selection]

14. Add the Selected Map attribute into the Filters card and select the Point Distribution checkbox in the filter dialogue.

Heatmap

1. Create a new sheet and name it Heatmap.
2. Add Longitude too the Columns card and Latitude to the Rows card. Right-click on each of the pills and select Dimension.
3. Set the Marks type to Density.
4. Drag-and-drop Country to Detail in the card.
5. Drag-and-drop Event Type to Detail in the card.

6. Set the mark Color to Density Dark Red and Intensity to 90%.
7. Lower the Size of the marks to about 1/4 of the bar.
8. On the Toolbar, change the map background color by selecting Map > Background Maps > Dark.

12. Show the Map Selection parameter and set the value to Heatmap.
13. Add the Selected Map attribute into the Filters card and select the Heatmap checkbox in the filter dialogue.

Line Chart

1. Create a new sheet called Line.
2. Create a new Parameter called Date Granularity with the following properties. Note: the values on the left column need to be in lower case.

3. Create a new Calculated Attribute called Date Truncated with the formula below:

DATE(DATETRUNC([Date Granularity], [Event Date]))

4. Create a new Parameter called Charts By with the following properties:

5. Create a new Calculated Attribute called Chart Count with the formula below:

IF [Charts By] = 'Events' 
THEN 1 
ELSE [Fatalities]
END

6. Add Date Truncated to the Columns card. Right-click on the pill and select Exact Date.
7. Add Chart Count to the Rows card.

8. Set the Marks type to Line.
9. Drag-and-drop Event Type to Color in the Marks card.
10. On the Toolbar, select Format > Shading. 11 Set the Worksheet shading color to #333333.

12. In the same format pane, switch to the Lines tab and set the Sheet Grid Lines to dotted line with the color #f5f5f5.

13. Set the titles of both axis to the color #f5f5f5.

14. Set the title of the worksheet to:

15. Create a Calculated Field named Date Display with the formular below:

CASE [Date Granularity]
    WHEN 'day' THEN STR([Date Truncated])
    WHEN 'week' THEN 'Week ' + DATENAME([Date Granularity],[Event Date]) + ' of ' + DATENAME('year', [Event Date])
    WHEN 'month' THEN DATENAME([Date Granularity],[Event Date]) + ' ' + DATENAME('year', [Event Date])
    WHEN 'quarter' THEN 'Q' + DATENAME([Date Granularity],[Event Date]) + ' ' + DATENAME('year', [Event Date])
    WHEN 'year' THEN DATENAME([Date Granularity],[Event Date])
END

16. Click on Tooltip in the Marks card and set it as follow:

Bar charts

1. Create a new sheet called Bar.
2. Create a new Calculated Attribute called Bar Size with the formula below. This formula is to calculate the width of a bar, with the 0.9 modifier used to create small gap between bars.

0.9*DATEDIFF('day', [Date Truncated], DATEADD([Date Granularity], 1, [Date Truncated]))

3. Add Date Truncated to the Columns card. Right-click on the pill and select Exact Date.
4. Add Chart Count to the Rows card.
5. Set the Marks type to Bar.
6. Drag-and-drop Bar Size to Size in the Marks card and change its measure to Average.
7. Click on Size in the Marks card and select Fixed. Set Alignment to Center.

8. Change the bar color to #f98b00.
9. On the Toolbar, select Format > Shading. 10 Set the Worksheet shading color to #333333.

12. Set the titles of both axis to the color #f5f5f5.
13. Set the title of the worksheet to:

Dashboard

Before we can start arranging the the dashboard, we need to create a couple more components.

1. Create a new Excel or csv file with the following data:

2. Create a new sheet in Tableau called Map Selection.
3. Add the excel (or csv) file into the sheet as a new data source by selecting Data > New Data Source on the Toolbar.
4. Create a new Parameter called Map Selection with the following properties:

5. Create a Calculated Attribute called Selected Map Color with the formula below:

IF [Chart Value]=[Map Selection]
    THEN 1
    ELSE 0
END

6. Add Chart Title to the Rows card.
7. Add Selected Map Color to Color in the Marks card.
8. Add Chart Title to Text in the Marks card.
9. Add Chart Value to Detail in the Marks card.
10. Set the sheet background color/shading to #333333 and remove all row and column lines.

11. Create a new sheet called Tooltip.
12. Add Chart Count to the Columns card and Event Type to the Rows card.
13. Right-click on Event Type and select Sort. Configure the sort as follow:

14. Add Event Type to Color in the Marks card.
15. Go back to the Bar worksheet.
16. Click on Tooltip in the Marks card and set it as follow:

17. Create a new Dashboard called Dashboard.
18. Set the dashboard Size to 1600x950.
19. Setup the initial layout for the dashboard as follow, using the Vertical object for the map, charts, and filter layout.

20. Set the Map Selection parameter to Point Distribution and add the Point worksheet into the Map Vertical Layout layout and remove the title.
21. Set the Map Selection parameter to Heatmap, which will make the Point sheet hidden on the dashboard, and add the Heatmap worksheet into the same Map Vertical Layout layout then remove the title.
22. Add the Line and Bar sheets into the Charts Vertical Layout layout, with the Line sheet on top.
23. Add all the filters into the Filters layout.
24. Remove the Map Selection parameter from the dashboard since we’re going to use a dashboard action to trigger the parameter.
25. Add the dashboard title and the map title worksheets into their respective layout. The dashboard should look like this at this point.

26. Add a new Vertical layout object into the Filter layout and add the Map Selection worksheet and the Event Type legend to it, with the Map Selection on top.

27. Make the new Vertical layout a Floating layout and set the background shading to #333333 and border to #f5f5f5.

28. Move the floating vertical layout to the top right of the map.

29. On the Toolbar, select Dashboard > **Action>>.
30. Add the following dashboard actions:

Final Dashboard

Link to Tableau Public

Alternative view with different map type and date granularity

Insights

• Most of the armed conflicts in South East Asia are centralized in Myanmar and Philippines.

• There is a large spike of events and fatalities during Q3 2016, which is due to the violence against civilians events in Philippines

• From 2015 to 2020, the most common armed conflict events in Myanmar are of the “Battles” type, with a larger cluster on the Western coast of the country.

• In recent years, protests are the most common type of armed conflicts in SOuth East Asia, with a relatively even split across many countries.

• “Battles” and “Violence against civilians” type of events are the most deadly, contributing to most of fatalities during the period from 2015 to 2020.