Annotate exponential function ggplot2 with variables

Question

I am trying to annotate the ggplot graph with the equation y=ae^(bx)

I fitted a linear model log(y) ~ x, and got the estimates for the coefficients: intercept and slope, and r^2 value:

a = round(exp(model$coefficients[1]), 2)
b = round(model$coefficients[2], 2)
r = round(summary(model)$r.squared, 2)

I would like to add the equation of the fitted model as an exponential to a ggplot graph, so lets assume my coefficients are:

a= -2
b= 0.5
r = 0.76

I used the example from: How to add the exponential regression equation and R2 on ggplot graph which uses the following code:

eq <- paste0('paste(y, " = ", italic(e^{',  round(exp.mod$coefficients[2], 2), 
             "*x ~~+~~ ", round(exp.mod$coefficients[1], 2),
             '}), ~~~~~~~~R^2~ "="~', round(summary(exp.mod)$r.squared, 2), ")")

but this gives a bit different formula (y=e^(ax+b)), I tried to rearrange it, but none of the attempts were successful

I also tried similar thing as shown here: Annotate exponential function ggplot2 but they used constants for their equation and I have variables, so I could not figure out how to replace numbers with variables

In the end, to add the equation to the plot, I tried to use the following line:

ggplot()+
annotate("text", aes(x=3, y=11), label = "y == a*e^{b*x}", parse =TRUE)

Pedro J. Aphalo · Accepted Answer · 2021-09-20T23:11:25.990

1

There are to problems with your last line of code: 1) annotate() does not support mapping of variables in data to aesthetics, one just the constant values as arguments; 2) the whole superscript, including the minus symbol must go inside the curly brackets. (The equation looks a bit nicer replacing the * in front of the exponential by ~ to add a bit of space, but this is optional.)

library(ggplot2)
ggplot()+
  annotate("text", x=3, y=11, label = "y == 13.048*e^{-0.697*x}", parse =TRUE)

Expressions have a tricky syntax. I used here your assumed values. I changed r into r.squared for clarity.

a= -2
b= 0.5
r.squared = 0.76

eq <- paste0('y~`=`~',  a,
             '~e^{', b , 
             '*x}*"; "*R^2~`=`~', r.squared)

ggplot()+
  annotate("text", x=3, y=11, label = eq, parse =TRUE)

^{Created on 2021-09-21 by the reprex package (v2.0.1)}

edited Sep 20 '21 at 23:11

answered Sep 20 '21 at 22:35

Pedro J. Aphalo

5,796
1
22
23

I guess you used a `log()` transformation as you tell that you fit a linear model. If you need to automate this for multiple plots you could also use package 'ggpmisc' (I am the author). – Pedro J. Aphalo Sep 20 '21 at 22:39
the issue is not adding numbers to equations: my coefficients are stored as a variable: eg a=13.048 from your example. this answer does not help to answer my question – yuliaUU Sep 20 '21 at 22:40
Please, tell the model formula you used in the fit. I think in your first block of code the call to `exp()` should for `a` instead of `b`. – Pedro J. Aphalo Sep 20 '21 at 22:50
@yuliaUU I edited my answer showing how to assemble the equation using the assumed variables. You can replace a, b and r.squared with code to extract the numbers from the model fit. – Pedro J. Aphalo Sep 20 '21 at 23:14
1

@yuliaUU The idea of Stackoverflow is that questions and answers remain available for future users, and that they are clearly written. That the code examples can be run, maybe demonstrating that an error is triggered is also important. Because of this, I edited your question. Keep in mind this in the future, please. – Pedro J. Aphalo Sep 20 '21 at 23:32

Annotate exponential function ggplot2 with variables

1 Answers1