how to combine two in one plot at r?

Question

data is fraway boot

I want to

Make a plot showing age- x axis and proportion of Down’s births in
each age-group -y axis. Add two curves showing the predicted probabilities
from the main effects model and the quadratic model. with legend.

lmod2 <- glm(prop ~ age + I(age^2),
             family=quasibinomial, finaldata)
lmod <- glm(prop ~ age,
            family=quasibinomial, finaldata)

things I tried:

plot(predict(lmod,lmod2, newdata = list(age = 16:50), type = 'response'), type = 'l',
+      ylab = "Probability of Down's syndrome per live birth", 
+      xlab = 'Maternal age')
Error in !se.fit : invalid argument type


> plot(lmod,lmod2)
Error in plot.lm(lmod, lmod2) : 'which' must be in 1:6

please guide me.

plot() will generate multiple plots (res vs fitted, Q-Q, etc.). Is there any specific plot you want out of your glm() models or will any plot be adequate? — jrcalabrese, Feb 12 '23 at 20:20
@jrcalabrese thanks for the guidence though they want the two curve in one plot with legend could you please guide? — fashionable, Feb 12 '23 at 20:26
Please provide enough code so others can better understand or reproduce the problem. — Community, Feb 12 '23 at 20:38
@jrcalabrese thanks. I had another question how can we fit a model HG as the response and the 3 covariates as main effects? Would this be the right way? m1 <- glm(cbind(HG,PI-HG)~ EH + NV, family = binomial, data = detectseparation::endometrial) — fashionable, Feb 12 '23 at 21:17
You can only include [one question per post](https://meta.stackexchange.com/a/222741/1285412). When you are able to post again, please make sure your post is [reproducible](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example). With my answer, I had to use `data(mtcars)` because you didn't provide `dput(finaldata)`. — jrcalabrese, Feb 12 '23 at 21:28

jrcalabrese · Accepted Answer · 2023-02-12T20:42:59.000

You can use par and mfrow to set the layout of what you want your plots to look like.

data(mtcars)
lmod <- glm(vs ~ mpg, family = quasibinomial, mtcars)
lmod2 <- glm(vs ~ mpg + I(mpg^2), family = quasibinomial, mtcars)

par(mfrow=c(1, 2))
# or par(mfrow=c(2, 1))
plot(predict(lmod, type = 'response'), type = 'l',
     ylab = "Probability of Down's syndrome per live birth", 
     xlab = 'Maternal age')
plot(predict(lmod2, type = 'response'), type = 'l',
     ylab = "Probability of Down's syndrome per live birth", 
     xlab = 'Maternal age')

If you would like both lines to be plotted on the same plot, you can change the second plot() to lines(). Just make sure to assign a different color.

data(mtcars)
lmod <- glm(vs ~ mpg, family = quasibinomial, mtcars)
lmod2 <- glm(vs ~ mpg + I(mpg^2), family = quasibinomial, mtcars)

par(mfrow=c(1,1))
plot(predict(lmod, type = 'response'), type = 'l', 
     col = "red",
     ylab = "Probability of Down's syndrome per live birth",  
     xlab = 'Maternal age')

lines(predict(lmod2, type = 'response'), type = 'l',
      col = "blue",
      ylab = "Probability of Down's syndrome per live birth", 
      xlab = 'Maternal age')

how to combine two in one plot at r?

1 Answers1