Skip to main content

Posts

Showing posts from 2012

The Arrow Charts

Quite recently, Jon Peltier suggested an interesting way of preparing Arrow Charts to show how a set of values changes between one point in time and another. The twist came in the manner of inserting shapes, in the form of suitable arrows, to replace the visible stacked bars in the chart. Inspired, I took the same data table and created my version of arrow charts with error bars.  The chart on the left is a Line chart, while the one to the right is a combination of Stacked bar and XY Charts. Since the budget values are represented in the numerical axes in both charts, I chose to include labels that would suggest the percentage change in budget for each division. 

The Dorling Cartogram

My last project involved using a multitude of regions for drawing analysis, parallels and comparison. Not wanting to use yet another Choropleth graph, I decided to look up alternatives that were easier to create and preferably required no VBA. Soon I stumbled upon "The Dorling Cartogram", defined in the UCSB site as, "This type of cartogram was named after its inventor, Danny Dorling of the University of Leeds. A Dorling cartogram maintains neither shape, topology nor object centroids, though it has proven to be a very effective cartogram method. To create a Dorling cartogram, instead of enlarging or shrinking the objects themselves, the cartographer will replace the objects with a uniform shape, usually a circle, of the appropriate size." I had the data for Obesity in the United States handy, so I decided to give it a try before using it in my project. I opted to use Bubble charts because data points within a series may need to be of varied shapes based on

The Pie-Doughnut Combination: A Radial Treemap

I stumbled into this graph some years ago, while looking up data for a project on US Auto Sales, through a post in Neoformix - Discovering and Illustrating Patterns in Data. So, when I started making the list of Pie-Doughnut combination charts, I decided to include this variety and use it to display survivors and victims of the Titanic tragedy in its centenary year.  It is not a particularly difficult chart to create, especially if one sets up the table properly. My first attempt brought me this: I thought a lot over whether to keep the statistics as labels for the outermost doughnut that displayed the survivor/victim proportions. They get a bit muddled up around the first-class women and children's section and the inability to add leader lines to the doughnut labels meant they couldn't be dragged away and the lines used as pointers. Here is the chart with the stats on display:

The Pie-Doughnut Combination: A Fan Plot

Happy to be back after a pretty busy beginning to the new year. I had this completed almost immediately following the preceding post but was otherwise hard-pressed to find a suitable time to post it. Soon after writing the Florence Nightingale Circumplex Chart post, I started searching for more varieties of charts that can be created using combinations of Pie-Doughnut charts. Soon enough, I found one through Naomi B. Robbins' comment in a Jorge Camoes' post . The referenced PDF article attempts to use Fan Plot to display relative quantities and differences using the R statistical language as is shown in the image below. Impassive of its benefits and/or disadvantages, I created it in Excel.

Florence Nightingale Circumplex Chart

I was taken to the  Florence Nightingale's Wiki page   during a recent research, and one of the interesting things I noted was her contribution to statistics. It came to me as a pleasant surprise that she is credited with inventing the polar area diagram , or occasionally the Nightingale rose diagram, which is equivalent to a modern circular histogram. Following the completion of my project and in my weekend to spare, I devoted time to recreating the chart in Excel. It took a combination of Doughnut-Pie-and XY charts and close to four hours to finish it. The colours are a bit darker, the values are approximate and the labels differently oriented, yet the chart looks fairly close to the original as is shown by the picture below.