Get the right hand side variables of an R formula

问题

I'm writing my first S3 class and associated methods and I would like to know how to subset my input data set in order to keep only the variables specified in the formula?

data(iris)
f <- Species~Petal.Length + Petal.Width

With model.frame(f,iris) I get a subset with all the variables in the formula. How to automatically keep only the right hand side variables (in the example Petal.Length and Petal.Width)?

回答1:

You want labels and terms; see ?labels, ?terms, and ?terms.object.

labels(terms(f))
# [1] "Petal.Length" "Petal.Width"

In particular, labels.terms returns the "term.labels" attribute of a terms object, which excludes the LHS variable.

回答2:

If you have a function in your formula, e.g., log, and want to subset the data frame based on the variables, you can use get_all_vars. This will ignore the function and extract the untransformed variables:

f2 <- Species ~ log(Petal.Length) + Petal.Width

get_all_vars(f2[-2], iris)

    Petal.Length Petal.Width
1            1.4         0.2
2            1.4         0.2
3            1.3         0.2
4            1.5         0.2
...

If you just want the variable names, all.vars is a very helpful function:

all.vars(f2[-2])

[1] "Petal.Length" "Petal.Width"

The [-2] is used to exclude the left hand side.

回答3:

One way is to use subsetting to remove the LHS from the formula. Then you can use model.frame on this:

f[-2]
~Petal.Length + Petal.Width

model.frame(f[-2],iris)
    Petal.Length Petal.Width
1            1.4         0.2
2            1.4         0.2
3            1.3         0.2
4            1.5         0.2
5            1.4         0.2
6            1.7         0.4
...

回答4:

The package formula.tools has a number of functions to make life easier working with formulas. In your case:

> formula.tools::rhs.vars(f)
[1] "Petal.Length" "Petal.Width"

Relying on base R can be dangerous because the left hand side can be missing, meaning that element 1 no longer refers to that.

来源：https://stackoverflow.com/questions/21330633/get-the-right-hand-side-variables-of-an-r-formula

标签

class

formula