Stewart Varner
Member
Hi everybody,
I am looking for a very powerful data visualization tool for historians. A professor has been teaching his students R in his Quantitative Methods class but (understandably) feels that the learning curve is too steep. He was asking me about creating a user interface for R but, before we think about how to go down that path, I thought I should check with the community.
So, any thoughts?
I suspect it might help if you could give us a bit more of a sense of the types of data to be visualized and the types of visualizations that are of interest. For static visualizations, I'd probably want to confirm that common charting tools in something like Excel or even more specialized charts in something like SPSS are really and truly insufficient. Those are obviously limited, but they tend to be easier to use and maybe more appropriate at some stages. Beyond that, it may be more specific tools for specific purposes, such as specialized timeline tools.
Stewart Varner
Member
As promised, I gathered some more information. here is what the professor is looking for:
"The size of the data set varies with the project. One student this year was working with thousands of observations from the Gallup poll, but another had only county-level data for West Virginia. I want to focus on visualizations that are poorly supported in other packages, in particular box plots (also known as box-whisker plots) and various variations such as notched plots, variable width plots, violin plots, and beam and fulcrum plots. Important for all of these is the order of the plots, and possible sub-groups of plots. Also underdeveloped are good ways of visualizing correlation matrices -- using color, shape, or size to mark the magnitude and direction of the correlation. And finally scatterplots."
Back when I was doing this type of analysis on a more regular basis, I found the Excel's statistical packages relatively robust without having to move to something like SPSS or SAS. However, since the goal is to work with the data in some more obscure output formats (or at least more computationally intensive outputs), you're a bit limited to packages that have a programmatic interface that allow the math analysis and programmatic, granular control of the output format.
That said, there are a lot of GUIs for R already which may get you further down this path (or at least allow you mask some of the complexity); one I thought was particularly well done (or at least reminded me more of MATLAB) was RStudio (http://www.rstudio.org/).
You must log in to post.
A project of the Association for Computers and the Humanities and the Chronicle of Higher Education's ProfHacker
Icons by famfamfam
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
