CODEWITHSUNDEEP
r
kittygirl
kittygirl

Reputation: 2443

How to wrap `factor level name` in `aggregate` function?

> levels(iris$Species)
[1] "setosa"     "versicolor" "virginica"

I plan to use above names as xlab in scripts as below: 

aggregate(Sepal.Length~Species,iris,function(x){
        car::qqPlot(x,distribution="norm",xlab="names(x)")
})

20200507002255276.jpg

After qqPlot,I found xlab not resolved.

The general questions is:
How to wrap factor level name in aggregate function like aggregate(y~factor,df,function(x){factor level name})?

Upvotes: 1

Views: 61

Answers (1)

Rui Barradas
Rui Barradas

Reputation: 76402

Here are two ways, I believe the question asks for the second one. None of th two needs aggregate.

1. The first way is to use the formula interface directly. It puts all graphs in the same page/plot.

car::qqPlot(Sepal.Length ~ Species, data = iris, distribution = "norm")

2. The second way splits the data set by the levels of the factor and then lapply's an anonymous function to plot each QQ plot separately.

sp <- split(iris, iris$Species)
lapply(seq_along(sp), function(i){
  sp[[i]]$Species <- droplevels(sp[[i]]$Species)
  car::qqPlot(Sepal.Length ~ Species, data = sp[[i]], 
              distribution = "norm", xlab = names(sp)[i])
})
rm(sp)    # final clean up

Upvotes: 1

Related Questions