How to call a concatenated vector of variable names in the right-hand side of a formula?

Question

How can I call a concatenated vector of variable names on the right-hand side of a formula?

For example, this works:

lm(data = mtcars, mpg ~ disp + hp)

But this does not:

rhs <- paste(c("disp", "hp"), collapse = " + ")
lm(data = mtcars, mpg ~ rhs)
#> Error in model.frame.default(formula = mpg ~ rhs, data = mtcars, drop.unused.levels = TRUE): variable lengths differ (found for 'rhs')

How can I avoid this error?

your rhs object is a string. Althggh it may print the same, you have to create a formula to evaluate within the function call. — GuedesBF, Sep 10 '21 at 00:46
Possible duplicate https://stackoverflow.com/questions/4951442/formula-with-dynamic-number-of-variables — Ronak Shah, Sep 10 '21 at 01:18

score 1 · Answer 1 · answered Sep 10 '21 at 00:45

by do.call and as.formula,

rhs <- paste("mpg ~", rhs)
rhs

"mpg ~ disp + hp"

do.call("lm", list(as.formula(rhs), data = as.name("mtcars")))

Call:
lm(formula = mpg ~ disp + hp, data = mtcars)

Coefficients:
(Intercept)         disp           hp  
   30.73590     -0.03035     -0.02484

GuedesBF · Accepted Answer · 2021-09-10T01:03:04.213

your rhs object is a string. If you have formulas stored as strings, you have to first convert to formula with as.formula, then use it as you would normally.

rhs <- paste(c("disp", "hp"), collapse = " + ")
tilde <- ' ~ '
lhs <- 'mpg ~ '

lm(data = mtcars, formula=as.formula(paste0(lhs, tilde, rhs)))

Call:
lm(formula = as.formula(paste0(lhs, tilde, rhs)), data = mtcars)

Coefficients:
(Intercept)         disp           hp  
   30.73590     -0.03035     -0.02484

As suggested in the comment by @rawr, you can also use reformulate:

lm(data = mtcars, formula=reformulate(termlabels = rhs, response = lhs))

How to call a concatenated vector of variable names in the right-hand side of a formula?

2 Answers2