Today I was reading about Ryan Sleepers tutorial on how to create Dumbbell charts. Coincidentally, an article popped up today on my twitter feed about "Why are US Presidential Elections So Close?" published by nautilus.com. Obviously what caught my attention is not the article itself but the chart that was put in the article.
The chart basically shows the % of popular votes between two prominent US political parties - Dems and Reps in the last 48 election seasons (ie from 1824 till 2012). The chart clearly articulates the subject of the topic - how big is the gap in the % of popular votes during each political season and clearly identifying where each party stands. Looking at the top circle tells us which party was the most favorite and the line joining the bubble gives a sense how wide the gap is between the parties. Simple yet very clear story.
Coincidentally, I have been reading a number of blogs on how to create dumbell / DNA / Barbell charts from here, here and here. With this triggering points, I thought why not I write my learnings about how to use charts with slight variations to tell variety of stories.....And thus led me to writing this blogspot.
Technically the charts look the same but the making is very different based on how the measures are used. The workbook at the end of this post, shows some variations of how a dumbbell chart is created in 2 scenarios -
1. When multiple measures are used - Corporate Sales and Home Office Sales
2. When same measure but cut by a dimension is used - Ex - Sales cut by Segment
It also depicts two of the use cases where they are apt to showcase -
1. Show how wide the gap is between the metrics - The following demonstrates how wide the gap is between the Home Office and Corproate Sales (Similar use case as what the Nautilus.com shown here). The charts are built based on the type of measures(single or multiple measures) used -
2. Show how the sales changes period over period across the dimension - Below chart shows how the sales have been YoY between two comparison years.
Self learning tip - Get to be bottom of Table calculations - what, why, how, where etc.....They are such a powerful calculations that if used smartly, we can do wonders with our vizzes. :)
Now moving on my next topic of Viz learnings.....
HDMO - Happy Data Make Over :)
The chart basically shows the % of popular votes between two prominent US political parties - Dems and Reps in the last 48 election seasons (ie from 1824 till 2012). The chart clearly articulates the subject of the topic - how big is the gap in the % of popular votes during each political season and clearly identifying where each party stands. Looking at the top circle tells us which party was the most favorite and the line joining the bubble gives a sense how wide the gap is between the parties. Simple yet very clear story.
Coincidentally, I have been reading a number of blogs on how to create dumbell / DNA / Barbell charts from here, here and here. With this triggering points, I thought why not I write my learnings about how to use charts with slight variations to tell variety of stories.....And thus led me to writing this blogspot.
Technically the charts look the same but the making is very different based on how the measures are used. The workbook at the end of this post, shows some variations of how a dumbbell chart is created in 2 scenarios -
1. When multiple measures are used - Corporate Sales and Home Office Sales
2. When same measure but cut by a dimension is used - Ex - Sales cut by Segment
It also depicts two of the use cases where they are apt to showcase -
1. Show how wide the gap is between the metrics - The following demonstrates how wide the gap is between the Home Office and Corproate Sales (Similar use case as what the Nautilus.com shown here). The charts are built based on the type of measures(single or multiple measures) used -
2. Show how the sales changes period over period across the dimension - Below chart shows how the sales have been YoY between two comparison years.
Self learning tip - Get to be bottom of Table calculations - what, why, how, where etc.....They are such a powerful calculations that if used smartly, we can do wonders with our vizzes. :)
Now moving on my next topic of Viz learnings.....
HDMO - Happy Data Make Over :)
