Comments on: Gapminder for Excel

By: Walter Jacques

Walter Jacques — Tue, 28 Apr 2009 20:19:28 +0000

Jon, nice work! I’ve wondered if the Gapminder software effects could be duplicated in Excel, and I thank you for showing that it can. The interpolation for smooting is brilliant! I looked at the chart and couldn’t figure out how you directly specified the data range for the chart, though. Also, I’d like to plot data for several states, which are in turn divided into ten different regions (so the regions would have different colored bubbles). Where in the data set would you accommodate for the different regions to discern between the bubble colors?

Thanks!

Walter

By: Jon Peltier

Jon Peltier — Thu, 16 Apr 2009 11:47:14 +0000

Tyler -

Thanks for the report. Sometimes my htaccess file is rewritten and the older permalinks are incorrectly resolved. I suspect it’s my host doing it. They’re ultra restrictive on what they allow me to do.

By: Tyler

Tyler — Wed, 15 Apr 2009 22:32:19 +0000

Hi Jon. The link for Gapminder for Excel II doesn’t seem to be working.

By: Jon Peltier

Jon Peltier — Mon, 08 Dec 2008 03:30:31 +0000

Eric - Great idea. Stay tuned this week for a new post named Gapminder For Excel II. Note: the post will be at that link in a couple days. I need to finish it up nicely.

By: Eric

Eric — Mon, 08 Dec 2008 01:07:01 +0000

Wonderful code. so simple and so clever.
I adopted it.

One thing is missing for me… the possibility to keep track of the way the bubble is moving… like a footprint.

any idea of how we can spot on the graph each interpolations (i mean not each but the main ones…)

By: Jon Peltier

Jon Peltier — Tue, 19 Aug 2008 17:52:57 +0000

Thanks Sarah, I’ll have to check that out. (I’ve fixed your link.)

By: Sarah

Sarah — Tue, 19 Aug 2008 15:35:34 +0000

I’m excited to see this–a while ago I created a Gapminder-type animated chart at the request of my boss, but it looks like you’ve implemented a couple of the features that I left out (interpolation to smooth the animation, a slider to change years, and changing bubble size). I look forward to trying them out. In my version, I added the option to trace the paths of some data points through time–easy enough to do, but in case you’re interested: Gender Parity.

By: Jon Peltier

Jon Peltier — Mon, 18 Aug 2008 14:21:30 +0000

A linear axis has the same spacing between values along its length, so that the same absolute change results in the same distance along an axis. For example, a change from 1000 to 2000 moves the marker as far as a change from 10,000 to 11,000.

A log scale expands the lower end of a scale and compresses the upper end. The result is that an equal relative change produces an equal distance along an axis. For example, a change from 1000 to 1100 moves the marker as far as a change from 10,000 to 11,000.

The log scale is often used where percentage changes are more important than absolute changes, such as pricing (stock pricing especially) and population studies.

By: Stéphane Nolf

Stéphane Nolf — Mon, 18 Aug 2008 12:56:44 +0000

Hi Jon,

First of all I have to say that I am definitively a fan of your work and I really reading you since about 2 months. Good job!

I had a small question regarding the presentation that supported the topic. During his show Hans Rosling used several time logarithmic scales. I know the option exists and is easy to use in Excel but what would be the correct use of it ? When do you think it is relevant to use such a scale in comparison with a linear scale ?

My first impresssion on some of the charts displayed by Hans Rosling was that something was wrong with scale and that “false” linearity was shown. It reminds me the chart that do not start at 0 but only show e.g. percentage between 80% and 100% and therefore exagerate the variation.

What do you think ?

Thanks !

By: derek

derek — Tue, 12 Aug 2008 15:01:04 +0000

I know, I’m not complaining.