[R] scatterplot function - double check: dashed lines

Jim Lemon jim at bitwrit.com.au
Wed Jun 9 10:59:26 CEST 2010


On 06/08/2010 08:44 PM, K F Pearce wrote:
> Hello everyone,
>
> This is just a quick double check.  It concerns the 'scatterplot function' in R.
>
>   I have 6 curves and I wish to represent each of them by a different kind of line (their colour must be black).
>
> The curves are derived from the cuminc function...the coordinates of which are in 'xx'.
>
> Upon reading the documentation in R, it looks like I can use the 'on/off' command for lty and I can merely run:
>
> plot(xx,color="black",lty=c("11","13","1343","73","2262","35"))
>
> according to the documentation, "13" (for example) means 1 'on' and 3 'off' .
>
> Does the above look OK ?
>
> Say, in another study, I wish to draw my 6 lines all in different colours (solid lines), I suppose that I could type:
>
> plot(xx, color=c("red","black","purple","yellow","brown","violet"), lty=1)
>
Hi Kim,
It depends upon what the object(s) are that contain the values for the 
curves. If these are in a matrix or data frame, you would probably want:

plot(xx[,1],type="l",lty="11")
lines(xx[,2],lty="13")

Although I agree with Greg, and find myself scratching my glabella in 
confusion when I have to decipher plots like that. If the lines are 
separated at the right edge of the plot, you might try leaving some 
extra margin there and labeling them as Greg suggested.

You can get a combination of these effects with something like this:

matplot(cbind(sort(rnorm(10)),sort(rnorm(10))),col=2:3,type="b")

Jim



More information about the R-help mailing list