R: Add legend to line plots in R without using ggplot2 or with using it but using my own legend?

Question

Consider the following code:

T=1
N=50
X=matrix(rnorm(N*4, mean = 0, sd = 1), 4, N)
t=seq(0,T,by=T/N)
t=head(t,-1)
ymax=max(X); ymin=min(X) #bounds for simulated prices

##Plot
led=c() # For ledgend purposes
ymax=max(X); ymin=min(X) #bounds for simulated prices
plot(t,X[1,], t='l', ylim=c(ymin, ymax), col=1, ylab="Price P(t)", xlab="time t")
led=c(led, paste("GBM", 1))
for(i in 2:4){
    lines(t, X[i,], t='l', ylim=c(ymin, ymax), col=i)
    led=c(led, paste("GBM", i))
}
legend(0, 0, legend=led, lty=1:4, cex=0.8)

The outcome would be

As you can see the legend is not observed and if they are observed they are of another colour of that of the lines.

How can I make the legend represent the colour without using the ggplot2? And how to use it with my own created legend?

Is that a requirement for some reason? ggplot2 has pretty good support for legends. — Jack Brookes
– Jack Brookes, Commented Aug 25, 2018 at 16:21
I'm confused about "without using the ggplot2" since you're not using ggplot. I'm also removing the jupyter-notebook tag since this isn't specific to those notebooks — camille
– camille, Commented Aug 25, 2018 at 17:38
@camille thank you but I do not understand why did you downgrade the question? I do want it to look good on jupyter notebook. As the drawings are not good looking there. — rsc05
– rsc05, Commented Aug 25, 2018 at 18:06

Jack Brookes · Accepted Answer · 2018-08-25 16:20:28Z

2

The code is incredible unreadable and is not done in a way you would normally plot in R. Here's a tidyverse solution that gets you to the same place.

tmax <- 1 # shouldnt use T as a variable name
N <- 50
X <- matrix(rnorm(N * 4, mean = 0, sd = 1), 4, N)
times <- seq(0, tmax, by = tmax / N) # shouldnt use t as a variable name
times <- head(times, -1)
ymax <- max(X)
ymin <- min(X) #bounds for simulated prices

library(tidyverse)

df <- as.data.frame(t(X)) %>% 
  mutate(time = times) %>% 
  gather(GBM, price, -time) %>% 
  mutate(GBM = gsub("V", "GBM ", GBM))

ggplot(df, aes(x = time, y = price, color = GBM)) +
  geom_line()

answered Aug 25, 2018 at 16:20

Jack Brookes

3,8402 gold badges15 silver badges24 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Jim Chen · Accepted Answer · 2018-08-25 16:18:30Z

1

Just add col argument in legend

T=1
N=50
X=matrix(rnorm(N*4, mean = 0, sd = 1), 4, N)
t=seq(0,T,by=T/N)
t=head(t,-1)
ymax=max(X); ymin=min(X) #bounds for simulated prices

##Plot
led=c() # For ledgend purposes
ymax=max(X); ymin=min(X) #bounds for simulated prices
plot(t,X[1,], t='l', ylim=c(ymin, ymax), col=1, ylab="Price P(t)", xlab="time t")
led=c(led, paste("GBM", 1))
for(i in 2:4){
  lines(t, X[i,], t='l', ylim=c(ymin, ymax), col=i)
  led=c(led, paste("GBM", i))
}
legend(0, 0, legend=led, cex=0.8,col = 1:4, lwd=2.5)

You can also change legend postion:

legend("topleft", legend=led, cex=0.8,col = 1:4,lwd=2.5)

edited Aug 25, 2018 at 16:18

answered Aug 25, 2018 at 16:09

Jim Chen

3,8694 gold badges24 silver badges39 bronze badges

3 Comments

rsc05 Over a year ago

Thank you so much for this but the legend is hiding some of the figure? Is there a neater way to display this? Furthermore, I see a dotted lines which is not the case in the figure. Is there a way to change that?

Jim Chen Over a year ago

The dot line is because of the argument "lty", just remove it.

Jim Chen Over a year ago

You can change the first argument of legend such as "topleft" "topright" "bottomleft" and so on to change the position of your legend.

Collectives™ on Stack Overflow

R: Add legend to line plots in R without using ggplot2 or with using it but using my own legend?

2 Answers 2

Comments

You can also change legend postion:

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

You can also change legend postion:

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related