trellis.device: Initializing Trellis Displays

Description

Initialization of a display device with appropriate graphical parameters.

Usage

trellis.device(device = getOption("device"),
               color = !(dev.name == "postscript"),
               theme = lattice.getOption("default.theme"),
               new = TRUE,
               retain = FALSE,
               ...)
standard.theme(name, color)
canonical.theme(name, color)
col.whitebg()

Arguments

device

function (or the name of one as a character string) that starts a device. Admissible values depend on the platform and how Rwas compiled (see Devices), but usually "pdf", "po

color

logical, whether the initial settings should be color or black and white. Defaults to FALSE for postscript devices, TRUE otherwise. Note that this only applies to the initial choice of colors, which can be overridde

theme

list of components that changes the settings of the device opened, or, a function that when called produces such a list. The function name can be supplied as a quoted string. These settings are only used to modify the default settings (deter

new

logical flag indicating whether a new device should be started. If FALSE, the options for the current device are changed to the defaults determined by the other arguments.

retain

logical. If TRUE and a setting for this device already exists, then that is used instead of the defaults for this device. By default, pre-existing settings are overwritten (and lost).

name

name of the device for which the setting is required, as returned by .Device

...

additional parameters to be passed to the device function, most commonly file for non-screen devices, as well as height, width, etc. See the help file for individual devices for admissible a

Value

standard.theme returns a list of components defining graphical parameter settings for Lattice displays. It is used internally in trellis.device, and can also be used as the theme argument to trellis.par.set, or even as theme in trellis.device to use the defaults for another device. canonical.theme is an alias for standard.theme.
col.whitebg returns a similar (but smaller) list that is suitable as the theme argument to trellis.device and trellis.par.set. It contains settings values which provide colors suitable for plotting on a white background. Note that the name col.whitebg is somewhat of a misnomer, since it actually sets the background to transparent rather than white.

Details

Trellis Graphics functions obtain the default values of various graphical parameters (colors, line types, fonts, etc.) from a customizable settings list. This functionality is analogous to par for standard Rgraphics and, together with lattice.options, mostly supplants it (par settings are mostly ignored by Lattice). Unlike par, Trellis settings can be controlled separately for each different device type (but not concurrently for different instances of the same device). standard.theme and col.whitebg produce predefined settings (a.k.a. themes), while trellis.device provides a high level interface to control which theme will be in effect when a new device is opened. trellis.device is called automatically when a "trellis" object is plotted, and the defaults can be used to provide sufficient control, so in a properly configured system it is rarely necessary for the user to call trellis.device explicitly.

The standard.theme function is intended to provide device specific settings (e.g. light colors on a grey background for screen devices, dark colors or black and white for print devices) which were used as defaults prior to R2.3.0. However, these defaults are not always appropriate, due to the variety of platforms and hardware settings on which Ris used, as well as the fact that a plot created on a particular device may be subsequently used in many different ways. For this reason, a safe default is used for all devices from R2.3.0 onwards. The old behaviour can be reinstated by setting standard.theme as the default theme argument, e.g. by putting options(lattice.theme = "standard.theme") in a startup script (see the entry for theme above for details).

References

Sarkar, Deepayan (2008) "Lattice: Multivariate Data Visualization with R", Springer. http://lmdvr.r-forge.r-project.org/