astsa (version 2.2)

tsplot: Time Series Plot

Description

Produces a nice plot of univariate or multiple time series in one easy line.

Usage

tsplot(x, y=NULL, main=NULL, ylab=NULL, xlab='Time', type=NULL,  margins=.25, 
        ncolm=1, byrow=TRUE, nx=NULL, ny=nx, minor=TRUE, nxm=2, nym=1, 
        xm.grid=TRUE, ym.grid=TRUE, col=1, gg=FALSE, spaghetti=FALSE, pch=NULL, 
        lty=1, lwd=1, mgpp=0, topper=NULL, addLegend=FALSE, location='topright', 
        boxit=TRUE, horiz=FALSE, legend=NULL, llwd=NULL, scale=1, ...)

Value

Produces a graphic and returns it invisibly so it can be saved in an R variable with the ability to replay it; see recordPlot.

Arguments

x, y

time series to be plotted; if both present, x will be the time index.

main

add a plot title - the default is no title.

ylab

y-axis label - the default is the name of the ts object.

xlab

x-axis label - the default is 'Time'.

type

type of plot - the default is line.

margins

inches to add (or subtract) to the margins. Input one value to apply to all margins or a vector of length 4 to add (or subtract) to the (bottom, left, top, right) margins.

ncolm

for multiple time series, the number of columns to plot.

byrow

for multiple time series - if TRUE (default), plot series row wise; if FALSE, plot series column wise.

nx, ny

number of major cells of the grid in x and y direction. When NULL, as per default, the grid aligns with the tick marks on the corresponding default axis (i.e., tickmarks as computed by axTicks). When NA, no grid lines are drawn in the corresponding direction.

minor, nxm, nym

if minor=TRUE, the number of minor tick marks on x-axis, y-axis. minor=FALSE removes both or set either to 0 or 1 to remove. The default is one minor tick on the x-axis and none on the y-axis.

xm.grid, ym.grid

if TRUE (default), adds grid lines at minor x-axis, y-axis ticks.

col

line color(s), can be a vector for multiple time series.

gg

if TRUE, will produce a gris-gris plot (gray graphic interior with white grid lines); the default is FALSE. The grammar of astsa is voodoo; see
https://www.youtube.com/watch?v=b4J8VrprrGE

spaghetti

if TRUE, will produce a spaghetti plot (all series on same plot).

pch

plot symbols (default is 1, circle); can be a vector for multiple plots.

lty

line type (default is 1, solid line); can be a vector for multiple plots.

lwd

line width (default is 1); can be a vector for multiple plots.

mgpp

this is used to adjust (add to) the mgp graphics parameters settings (?par), which are c(1.6,.6,0) here; the R default is c(3,1,0). This will be helpful in moving an axis label farther from the axis if necessary.

topper

non-negative value to add to the top outer margin; if NULL (default) a suitable value is chosen

addLegend

if TRUE and spaghetti=TRUE, will add a simple legend. If more details are needed, leave this set to FALSE (the default) and use legend directly.

location

if addLegend=TRUE, the location of the legend with options "bottomright", "bottom", "bottomleft", "left", "topleft", "top", "topright" (the default), "right" and "center".

boxit

if TRUE (default), the legend is in a box; if FALSE, no box is drawn.

horiz

if addLegend=TRUE, should the legend be horizontal (default is FALSE - vertical).

legend

if NULL (default), the legend uses names of each time series; otherwise, use to customize legend.

llwd

line width for the legend if different from the plotted lines.

scale

for multiple series, the scale for character expansion (cex = .9*scale).

...

other graphical parameters; see par.

Author

D.S. Stoffer

References

You can find demonstrations of astsa capabilities at FUN WITH ASTSA.

The most recent version of the package can be found at https://github.com/nickpoison/astsa/.

In addition, the News and ChangeLog files are at https://github.com/nickpoison/astsa/blob/master/NEWS.md.

The webpages for the texts and some help on using R for time series analysis can be found at https://nickpoison.github.io/.

Examples

if (FALSE) {

# minimal
tsplot(soi)
# prettified
tsplot(soi, col=4, main="Southern Oscillation Index")

# gris-gris multiple plot 
tsplot(climhyd, ncolm=2, gg=TRUE, col=2:7, lwd=2)  

# spaghetti (and store it in an object - ?recordPlot for details)
x <- replicate(100, cumsum(rcauchy(1000))/1:1000)
u <- tsplot(x, col=1:8, main='No LLN For You', spaghetti=TRUE)
u   #  plot on demand

# quick legend with spaghetti
tsplot(cbind(Mortality=cmort, Pollution=part), col=5:6, gg=TRUE, spaghetti=TRUE, addLegend=TRUE)
# ... and a little more control
tsplot(cbind(gtemp_land, gtemp_ocean), spaghetti=TRUE, lwd=2, col=astsa.col(c(4,2),.7), 
  ylab="\u00B0C", main="Global Temperature Anomalies", addLegend=TRUE, location='topleft', 
  horiz=TRUE, legend=c("Land Surface","Sea Surface"))


# compare these
par(mfrow=2:1)
tsplot(1:453, soi, ylab='SOI', xlab='Month')
# now recklessly add to the margins and add to mgp to get to the default
tsplot(1:453, soi, ylab='SOI', xlab='Month', margins=c(2,3,4,5), las=1, mgpp=c(1.4,.4,0))


# cex (and scale)
par(mfrow=c(3,1))  # cex gets small with mf plots 
  tsplot(cmort); tsplot(tempr); tsplot(part)
par(mfrow=c(3,1), cex=.9)  # so fix it 
  tsplot(cmort); tsplot(tempr); tsplot(part)

# it's rescaled here for your pleasure
tsplot(cbind(big=rnorm(100), bad=rnorm(100), john=rnorm(100)))  # default scale=1
tsplot(cbind(big=rnorm(100), bad=rnorm(100), john=rnorm(100)), scale=1.5)  # big
}