| denstrip-package {denstrip} | R Documentation |
Overview of the denstrip package
Description
Graphical methods for compactly illustrating and comparing distributions, particularly distributions arising from parameter estimation or prediction.
Details
denstrip implements the density strip for
illustrating a single univariate distribution. The darkness of the
density strip at a point is proportional to the density at that point.
A shortcut function denstrip.normal draws the strip for the
given normal distribution.
densregion implements the density region, which
illustrates the uncertainty surrounding a continuously-varying quantity
as a two-dimensional shaded region with darkness proportional to the
density. There are shortcut functions densregion.normal
and densregion.survfit for computing and drawing the
region for normally-distributed predictions and survival curves,
respectively.
sectioned.density implements the sectioned density plots
of Cohen and Cohen (2006). These illustrate distributions using
occlusion and varying shading. They were developed for
summarising data, but can also be used for illustrating known
distributions.
vwstrip can be used to draw varying-width strips to
illustrate distributions, in a similar manner to the violin plot for
summarising data. The width of the strip is proportional to the density.
A shortcut function vwstrip.normal draws the strip for the
given normal distribution.
bpstrip adapts the box-percentile plot to illustrate a
distribution instead of observed data. This strip has width
proportional to the probability of a more extreme point.
cistrip implements the popular point and line figure for
illustrating point and interval estimates, for example from multiple
regression.
These methods are discussed in more detail by Jackson (2008).
Each function is designed to add a graphic to an existing set of plot axes. The plots can be added to either base graphics or lattice panels.
Author(s)
Christopher Jackson <chris.jackson@mrc-bsu.cam.ac.uk>
References
Jackson, C. H. (2008) Displaying uncertainty with shading. The American Statistician, 62(4):340-347.