Discussion: View Thread

An interesting R package illustrating the design of complex graphs

  • 1.  An interesting R package illustrating the design of complex graphs

    Posted 07-15-2014 18:02
    Hi everyone,

    For those of you who work with R for statistical analyses and reporting, I wanted to mention an interesting R package I discovered while attending the recent UseR! conference which took place at UCLA.

    The package is named agridat and was designed by Kevin Wright.  The package includes a variety of agricultural data sets and uses these data sets to create complex and beautiful graphs. (Some of you may be interested in the actual data sets and/or the statistical modeling techniques used to analyze these data sets.)

    The agridat package vignette is available at the following website:
     
    http://cran.r-project.org/web/packages/agridat/vignettes/agridat.pdf

    The agridat package website is available at
    http://cran.r-project.org/web/packages/agridat/

    If you have any questions  or comments about the package, you can contact Kevin by e-mail at Kevin Wright <kw.stat@gmail.com" shape="rect" rel="nofollow" target="_blank">kw.stat@gmail.com>.

    To install the agridat package in R, you can use the R command: 

     install.packages("agridat")

    To load the agridat package into your current R session, you can use the command:

    require(agridat)

    To view the agridat package vignette from within R, you can use the command:

    vignette("agridat") 

    Kevin Wright was kind enough to share some quotes on designing statistical graphics with me and I am reproducing them below for your enjoyment: 

    "Good statistical graphics are...much harder than running regressions and making tables."
      Andrew
    Gelman (2011), J Comp Graph Stat, 20, 36-40.


    "
    A big challenge is balancing the tension between exploration and presentation. For exploratory graphics, you want to spend five seconds (or less) to create a plot that helps you understand the data, while you might spend five hours on a plot that's persuasive to an audience who isn't as intimately familiar with the data as you.''

      Hadley Wickham, http://simplystatistics.tumblr.com/post/22844703875/ha



    "The sweat is in the details....
    Most graphs, like many other endeavors follow the 80-20 Rule: You can get it 80% done with 20% of the effort, but the remaining 20% is hard work, and takes the remaining 80% of the effort.''

      Michael Friendly, http://www.datavis.ca/gallery/excellence.php



    "
    A little effort went into splitting the data set into cells, some effort went into making a rudimentary function to plot the data...and a great deal of effort went into attending to details''.

      Dan Carr (1995)


    "We
    made countless tweaks to our R script and hard-coded various details such as the positions of the labels, the scaling of the axes, the long tick mark dividing 2009 from 2010, and a slight shifting of the zero points so they would be clear of the x-axis (which we wanted to be precisely at zero - contrary to the R default - so make it visually apparent that the estimated rate declines to zero at the end of the time series). Some manual adjustments may always be necessary to prepare a complex presentation-quality graph, but R (as currently configured) is not the ideal environment for this process."
     
    Niemi & Gelman
    (2011), Making information clear-and beautiful. 
     
    Significance, 8, 135-137.



    Kind regards,

    Isabella

    -------------------------------------------
    Isabella Ghement
    Ghement Statistical Consulting Company Ltd.
    -------------------------------------------