4

Could someone tell me how to add a text line to lattice plot.

My code is: 

xyplot(Neff~Eeff,data=phuong,panel=mypanel,
       col="black",
       pch=18,xlab="Energy efficiency (%)",
       ylab = "Nitrogen efficiency (%)", main="(a)")

Here pane=mypanel is to add an abline. I want to add a linear regression equation between Eeff and Neff to their plot.

Thanks!

Mihai Labo
  • 1,082
  • 2
  • 16
  • 40
hn.phuong
  • 835
  • 6
  • 15
  • 24
  • 2
    See `?panel.text`, a function which you can add to your `mypanel` function. – BenBarnes Sep 03 '12 at 13:35
  • It does not work for me. This is mypanel codes: mypanel<-function(x,y,...){ + panel.xyplot(...) + panel.abline(lm(Neff~Eeff,data=phuong)) + panel.text(30,30,labels=tp) + } Could you please help me to see it? – hn.phuong Sep 03 '12 at 14:15
  • Please post a [reproducible example](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) so that it is easier to help you. – BenBarnes Sep 03 '12 at 14:18
  • Thank you very much for your comments Benbarnes. My problem is in fact like this. As you can see in my codes above, I have two collumes of data: Eeff and Neff 1. I made a scatter plot using xyplot function 2. I used simple linear regression function: lm(Eeff~Neff, data=phuong) to find the regression equation: which is Eff=19.7 + 0.36 Neff NOW, what I want is to add "Eff=19.7 + 0.36 Neff" to the top-left corner of the scatter plot. Could you please help me? Thank you very much! – hn.phuong Sep 03 '12 at 14:25
  • @hn.phuong this is still not reproducible. Ben provided a link. Check out his link. The basic idea is that in your original post we should be able to grab exactly the code you used and reproduce your error. You will want the data set to be as small as possible (maybe 5-10 rows) and still be able to reproduce the outcome. You can also use one of R's many built in data sets as an example. – Tyler Rinker Sep 03 '12 at 16:01
  • Thank you very much Tyler Rinker. Actually, I did not really understand how "reproducible data" should be yesterday. But now I understand, so it is good if I ask for helps next time. – hn.phuong Sep 04 '12 at 08:26
  • Thank BenBarnes very much! It works well and I have had what I want! – hn.phuong Sep 04 '12 at 08:27

2 Answers2

9

So, with reproducible data:

set.seed(1)

phuong <- data.frame(Eeff = rnorm(100,31))
phuong$Neff <- rnorm(100,19.7 + .36*phuong$Eeff)

# Create the lm object ahead of time
lm1 <- lm(Neff ~ Eeff, data = phuong)

# Create the character string that you want to print
tp <- sprintf("%s=%.1f + %.2f %s", all.vars(formula(lm1))[1],
  coef(lm1)[1], coef(lm1)[2], all.vars(formula(lm1))[2])

# Change the mypanel function to use the lm1 object
mypanel<-function(x,y,...){
  panel.xyplot(x, y, ...)
  panel.abline(lm1)
  panel.text(30,33,labels=tp)
}

library(lattice)

# and off we go.

xyplot(Neff~Eeff,data=phuong,panel=mypanel,
       col="black",
       pch=18,xlab="Energy efficiency (%)",
       ylab = "Nitrogen efficiency (%)", main="(a)")

enter image description here

BenBarnes
  • 19,114
  • 6
  • 56
  • 74
  • Hi BenBarnes, thank you very much for this help. I have an another small question related to this. This case is to add the linear equation of these two variable. What should I do i f I want to add a random words, e.g. Thanks! – hn.phuong Sep 04 '12 at 08:42
  • Hi @hn.phuong, to add other text, modify the `tp` object if you've already created it, for example using `paste(tp, "Thanks!")`. Or if you just want to add "Thanks!" without the regression equation, then `tp <- "Thanks!"`. Then create the scatterplot again using `xyplot`. – BenBarnes Sep 04 '12 at 08:48
  • @hn.phuong, if this answer solves your problem, please consider marking it as the accepted answer (check mark near the beginning of the answer). – BenBarnes Sep 04 '12 at 09:19
5
library(latticeExtra)

set.seed(1)

phuong <- data.frame(Eeff = rnorm(100,31))
phuong$Neff <- rnorm(100,19.7 + .36*phuong$Eeff)

xyplot(Neff ~ Eeff, data= phuong) + 
  layer(panel.ablineq(lm(Neff ~ Eeff, data= phuong), r.sq= TRUE, rot = TRUE))

enter image description here

user2030503
  • 3,064
  • 2
  • 36
  • 53