Adjusting the limits of x and y axis, when adding new curves to a plot in R

Question

I have two datasets (df1 and df2) that are plotted.

df1 = data.frame(x=c(1:10), y=c(1:10))
df2 = data.frame(x=c(0:13), y=c(0:13)^1.2)

# plot
plot(df1)
# add lines of another dataset
lines(df2)

Some values of df2 are out of the plot-range and thus not visible. (In this example I could just plot df2 first). I usually try to find out the ranges of my data, as shown below.

# manual solution
minX = min(df1$x, df2$x)
minY = min(df1$y, df2$y)
maxX = max(df1$x, df2$x)
maxY = max(df1$y, df2$y)

plot (df1, xlim=c(minX, maxX), ylim=c(minY, maxY))
lines(df2)

When having many datasets, this becomes annoying. I was wondering, if there is an easier way of adjusting the ranges of the axis. In the first step R finds axis ranges itself. Is there also a way that R adjusts the axis-ranges, when new datasets are added?

Answer 1

You could use range to calculate the limits.

Imho, a better solution:

df1 <- data.frame(x=c(1:10), y=c(1:10))
df2 <- data.frame(x=c(0:13), y=c(0:13)^1.2)

ll <- list(df1,df2)

ll <- lapply(1:length(ll),function(i) {res <- ll[[i]]; res$grp <- i; res})

df <- do.call("rbind",ll)
df$grp <- factor(df$grp)

library(ggplot2)
p1 <- ggplot(df,aes(x=x,y=y,group=grp,col=grp)) + geom_line()
p1

Answer 2

I like @Roland's solution, but here is an extension of @Glen_b's solution that works for an arbitrary number of data sets, if you have them all in a list.

(warning: untested!)

dflist <- list(df1,df2,df3,...)  ## dots are not literal!
plotline <- function(L,...) {    ## here the dots are literal
  ## use them to specify (e.g.) xlab, ylab, other arguments to plot()
  allX <- unlist(lapply(L,"[[","x"))
  allY <- unlist(lapply(L,"[[","y"))
  plot (df1, xlim=range(allX), ylim=range(allY),type="n",...)
  invisible(lapply(L,lines))
}

This assumes that you want all the data sets drawn as lines. If you want to start specify separate colours, point types, etc., you could extend this function -- but you would be starting to re-invent the lattice and ggplot2 packages at that point.

(If all your data sets are the same size, you should consider matplot)