These are panel functions for Trellis displays useful when a grouping
variable is specified for use within panels. The x
(and
y
where appropriate) variables are plotted with different
graphical parameters for each distinct value of the grouping variable.
panel.superpose(x, y = NULL, subscripts, groups,
panel.groups = "panel.xyplot",
…,
col, col.line, col.symbol,
pch, cex, fill, font,
fontface, fontfamily,
lty, lwd, alpha,
type = "p", grid = FALSE,
distribute.type = FALSE)
panel.superpose.2(…, distribute.type = TRUE)panel.superpose.plain(…,
col, col.line, col.symbol,
pch, cex, fill, font,
fontface, fontfamily,
lty, lwd, alpha)
Coordinates of the points to be displayed. Usually numeric.
The panel function to be used for each subgroup of points. Defaults
to panel.xyplot
.
To be able to distinguish between different levels of the
originating group inside panel.groups
, it will be supplied
two special arguments called group.number
and
group.value
which will hold the numeric code and factor level
corresponding to the current level of groups
. No special
care needs to be taken when writing a panel.groups
function
if this feature is not used.
An integer vector of subscripts giving indices of the x
and
y
values in the original data source. See the corresponding
entry in xyplot
for details.
A grouping variable. Different graphical parameters will be used to
plot the subsets of observations given by each distinct value of
groups
. The default graphical parameters are obtained from
the "superpose.symbol"
and "superpose.line"
settings
using trellis.par.get
wherever appropriate.
Usually a character vector specifying how each group
should be drawn. Formally, it is passed on to the
panel.groups
function, which must know what to do with it.
By default, panel.groups
is panel.xyplot
, whose
help page describes the admissible values.
The functions panel.superpose
and panel.superpose.2
differ only in the default value of distribute.type
, which
controls the way the type
argument is interpreted. If
distribute.type = FALSE
, then the interpretation is the same
as for panel.xyplot
for each of the unique groups. In other
words, if type
is a vector, all the individual components are
honoured concurrently. If distribute.type = TRUE
,
type
is replicated to be as long as the number of unique
values in groups
, and one component used for the points
corresponding to the each different group. Even in this case, it is
possible to request multiple types per group, specifying type
as a list, each component being the desired type
vector for
the corresponding group.
If distribute.type = FALSE
, any occurrence of "g"
in
type
causes a grid to be drawn, and all such occurrences are
removed before type
is passed on to panel.groups
.
Logical flag specifying whether a background reference
grid should be drawn. See panel.xyplot
for details.
A vector color specification. See Details.
A vector color specification. See Details.
A vector color specification. See Details.
A vector plotting character specification. See Details.
A vector size factor specification. See Details.
A vector fill color specification. See Details.
A vector color specification. See Details.
A vector color specification. See Details.
A vector color specification. See Details.
A vector alpha-transparency specification. See Details.
Extra arguments. Passed down to panel.superpose
from panel.superpose.2
, and to panel.groups
from
panel.superpose
.
logical controlling interpretation of the
type
argument.
panel.superpose
divides up the x
(and optionally
y
) variable(s) by the unique values of
groups[subscripts]
, and plots each subset with different
graphical parameters. The graphical parameters (col.symbol
,
pch
, etc.) are usually supplied as suitable atomic vectors, but
can also be lists. When panel.groups
is called for the
i
-th level of groups
, the corresponding element of each
graphical parameter is passed to it. In the list form, the individual
components can themselves be vectors.
The actual plot for each subgroup is created by the
panel.groups
function. With the default panel.groups
,
the col
argument is overridden by col.line
and
col.symbol
for lines and points respectively, which default to
the "superpose.line"
and "superpose.symbol"
settings.
However, col
will still be supplied as an argument to
panel.groups
functions that make use of it, with a default of
"black"
. The defaults of other graphical parameters are also
taken from the "superpose.line"
and "superpose.symbol"
settings as appropriate. The alpha
parameter takes it default
from the "superpose.line"
setting.
panel.superpose
and panel.superpose.2
differ essentially
in how type
is interpreted by default. The default behaviour
in panel.superpose
is the opposite of that in S, which is the
same as that of panel.superpose.2
.
panel.superpose.plain
is the same as panel.superpose
,
except that the default settings for the style arguments are the
same for all groups and are taken from the default plot style.
It is used in xyplot.ts
.
Different functions when used as panel.groups
gives different
types of plots, for example panel.xyplot
,
panel.dotplot
and panel.average
(This can
be used to produce interaction plots).
See Lattice
for an overview of the package, and
xyplot
for common arguments (in particular, the
discussion of the extended formula interface and the groups
argument).