Adding a stat_function layer to a geom_point in ggplot2

Question

I am trying to make a trajectory line representing brain atrophy over time. I have two specific trajectories for two different groups, and I want to add each of these two lines (for comparison) to a scatterplot of Age by Brain Volume.

The y-intercepts of these lines are 565 and 286, yet using the stat_function layer it seems to be placing the intercepts at very weird values, corrupting the overall scatterplot. Of course I can use ylim(), but then the functions don't appear (and they'd still be mis-specified even if they did).

My code:

ggplot(na.omit(GRAPHICSMATRIX), aes(x=Age, y=FrontalPole, colour=Dx_Bl)) +
  geom_point(size=3) +  
  ylim(0, 500) + 
  stat_function(fun=function(x) {-2.129*x^3 - 23.456*x + 565}, colour="blue", size=1) +  
  stat_function(fun=function(x) (286)+((-0.997)*(x)^3), colour="darkred", size=2)

Here are the graphs.

Correct Scatterplot

Incorrect Scatterplot with two stat_function()

Did you run function(x) {-2.129*x^3 - 23.456*x + 565} with GRAPHICSMATRIX$Age as input outside of the ggplot2 code? — hrbrmstr
– hrbrmstr, Commented May 29, 2016 at 2:25
The y-intercepts are reasonable numbers, but those are evaluated at x=0. Your plots start around x=60 in which case you function gives a result on the order of 216,000. Just plug in x=60 and you can see that. — Mike H.
– Mike H., Commented May 29, 2016 at 3:27

Mike H. · Accepted Answer · 2016-05-29 03:33:08Z

2

Yes, your y-intercepts are 565 and 286 respectively, but those values are for when x=0. The plots you show in your question start at around x=55. In this case when we plug in x=55 for your functions we get:

-2.129(55^3) - 23.456(55) + 565 = 354,937.455

286 + -0.997*(55^3) = 165,589.875

So what you're thinking is your y-intercept is not actually. I think you stat_functions are working correctly. If you don't like the results I think you need to re-think the functions you want to plot.

answered May 29, 2016 at 3:33

Mike H.

14.4k2 gold badges33 silver badges39 bronze badges

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

1 Comment

Brandon Klinedinst Over a year ago

Thank you. My model was based on a demographic of 55-90 year olds, but the model calculations presume an intercept of 0. So I solved this issue by shifting the functions to the right by 55, which I did by subtracting 55 at each X (ie (x-55)^2) Thank you!

Collectives™ on Stack Overflow

Adding a stat_function layer to a geom_point in ggplot2

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related