Home > Software engineering >  Plot multiple columns of points in ggplot
Plot multiple columns of points in ggplot

Time:02-21

I have a data frame structured as follows:

df <- data.frame(combo = c("first", "second", "last"),
                   effect.size = cumsum(rnorm(3)),
                   upper.CI = cumsum(rnorm(3, mean = 3, sd=1)),
                   lower.CI = cumsum(rnorm(3, mean = -3, sd=1)))

I want to plot each combo with their associated confidence intervals. However, I am not sure how to plot extra points and connect them on the same line. The intended output is something like: enter image description here

I am currently here, but can't figure out how to add the CI's:

ggplot(df, aes(x=effect.size, y=combo)) 
  geom_point() 
  geom_vline(xintercept = 0, linetype = 2) 
theme_classic2()

All of the values I am trying to plot are singular numbers that are an output from a test, so I don't have a distribution of numbers to calculate CI's within ggplot.

Any help is appreciated.

CodePudding user response:

You can use geom_pointrange or geom_linerange if you like.

ggplot(df, aes(effect.size, y=combo))   
  geom_pointrange(aes(xmin=lower.CI, xmax=upper.CI))  
  geom_vline(aes(xintercept = 0),linetype="dashed")
  • Related