ts.plot(..., abb = T, type = "l", lty = 1:8, pch = 0:18, col = 1:4, xlab = <<see below>>, ylab = "", xlim = <<see below>>, ylim = <<see below>>) ts.lines(..., type = "l", lty = 1:8, pch = 0:18, col = 1:4) ts.points(..., type = "p", lty = 1:8, pch = 0:18, col = 1:4)
TRUE
, labels for times series with
calendar time domains are abbreviated.
{"p", "l", "b", "o", "n", "h"}
,
telling which type of
plot (points, lines, both, overlaid, none or high-density) should be done for
each channel. The first character of type defines the first plot, the second
character the second, etc. Characters in type are used cyclically;
e.g.,
"pl"
alternately plots points and lines. Default is
"l"
for
ts.plot
and
ts.lines
,
"p"
for
ts.points
.
1:8
. Some devices have fewer than eight line types.
"p"
, the second for the second, etc. Default is 0:18.
To mix special plotting symbols (those specified by integers) and
plotting characters, either use character escapes, for example,
pch="X 02"
, or figure out the numeric equivalent of the ASCII
character after current, for example,
pch=c(88,2)
.
1:4
.
ts.plot
only). The default is to use the
time units of the series, if any, otherwise the string
"time"
. Use
""
if no label is desired.
ts.plot
only). Default is to leave it unlabeled.
ts.plot
only). Warnings are printed if values exceed these limits.
Default is the
min
and
max
of the union of the time values for all series.
ts.plot
only). Warnings are printed if values exceed these limits.
Default is the
min
and
max
of the union of the observation values for all
series.
ts.plot
function generates a new graphic with the values of each
series or component sub-series plotted against the observation time;
ts.points
and
ts.lines
add to the current graphic.
When the list of time-series arguments supplied to
ts.plot
includes both
regular and calendar time series, the calendar time series are treated as
regular (class
"rts"
), using the function
as.rts
to do the conversion.
No conversion is applied to irregular calendar time series.
Since
ts.lines
and
ts.points
have no way of knowing what types of
series are supplied to
ts.points
, it is possible that the result of calling
ts.plot
followed by
ts.lines
and/or
ts.points
might be different from
what would be obtained if all the arguments were given initially to
ts.plot
.
Other high-level graphical parameters may also be supplied as arguments
to this function, in particular the arguments to
title
(see
par
).
Factors are ignored in multivariate time series; in univariate time
series they are refused.
Because plotting symbols are drawn with lines and because these functions may
be changing the line style, you should probably specify
lty=1
when using
plotting symbols.