Metric_Single_Variable_Analysis: Single Variable Analysis of a metric variable

In corneliusfritz/intervisu: Interactive data visualisation with shiny

Description

With this application you can graphically analyse a single metric variable. From a given data-set you can plots boxplots, density-estimations and histograms, each with many interactive elements. If you want to reproduce the same graph you built in the app, you can always get the code for the desired representation by clicking the 'Show R Code'-Button.

Usage

Metric_Single_Variable_Analysis(data, n = 10, a = 50, width = 700,
  height = 700)

Arguments

data	A data.frame object that is to be analyzed (only metric variables will be used in this application)
n	A numeric value indicating the limit from what number of different values a variable is seen as categorical variable, all variables that have more than n different values are being treated as metric values
a	A numeric value indicating how much freedom the slider telling the boxplot what xlim values should be allowed should have
width	A numeric value indicating the width of the shown plot
height	A numeric value indicating the height of the shown plot

Details

There are several interactive elements to each interactive data visualizations. If you want to plot a boxplot you can at first decide what metric variable of the chosen data should be plotted. You can also plot the points in the boxplot by pressing the button named 'Show Points'. In order to see all observed data a jitter function is used to randomize the points in the x-coordinates. Another possibility is to show a horizontal boxplot. With two sliders you can also decide what values of the observed variable should be used and what values should be ploted in the boxplot. In order to see the R-code leading to the desired graphical presentation check the 'Show R Code' button. You can copy and paste the code in you R console.

Plotting a density estimator also gives you many different possibilities. At first you can decide what metric variable should be plotted. You can also change the used kernel, the options are a gaussian, epanechnikov, triangular, rectangular, cosine and opt-cosine kernel. With two numeric inputs you can decide from what value to what value the slider input for the chosen bandwidth should be shown. You can also just plot only the observed values of a variable that lie within a given range (with a slider you can decide what values should be used).

The last optional graphical representation is a histogram of the relative frequency. With a numeric input you can set the upper limit of the y-axis. You can also change the origin of the histogram. The origin of a histogram is where the first block of the histogram starts, and to plot all optional values it can't be greater than the smallest observed value. The last interactive parameter is the used bandwidth in the plot, this is the absolute length of each single bars in the histogram.

Author(s)

Cornelius Fritz <[e-mail]>

Examples

data=mtcars
if (interactive()) {
  Metric_Single_Variable_Analysis(data)
  }

corneliusfritz/intervisu documentation built on May 13, 2019, 10:51 p.m.