Passing formula as a quoted string to regsubsets() function from the leaps package

Question

I was trying to shorten the process of passing the formula to the regsubsets() function instead of having to write the full string. So, I used the following code for generation of formula string but the regsubsets function gave the error "argument "y" is missing".

When I pasted the generated formula string without quotes, it was accepted. But when I pasted formula string within quotes, the same error was generated. So, it seems to me that, the quotes are the problem.

How can I bypass this error?
Is there another function that can pass a string argument without quotes to such picky functions?

Here's the code sample:

# load data
data(mtcars)  # pre-loaded in R
head(mtcars,2)
#               mpg cyl disp  hp drat    wt  qsec vs am gear carb
# Mazda RX4      21   6  160 110  3.9 2.620 16.46  0  1    4    4
# Mazda RX4 Wag  21   6  160 110  3.9 2.875 17.02  0  1    4    4

## full model formula
# response & predictors
yvar = c('mpg')
xvars = setdiff(colnames(mtcars), yvar)
xvars
# 'cyl''disp''hp''drat''wt''qsec''vs''am''gear''carb'

# formula string
fstr = paste(yvar, '~', paste(setdiff(names(mtcars), yvar), collapse='+'))
fstr
# 'mpg ~ cyl+disp+hp+drat+wt+qsec+vs+am+gear+carb'

## regsubsets regression
library(leaps)
subset = regsubsets(fstr, data=mtcars, method='exhaustive', nbest=2)
res = summary(subset)
res

# Error in leaps.setup(x, y, wt = weights, nbest = nbest, nvmax = nvmax, : argument "y" is 
# missing, with no default
# Traceback:

# 1. regsubsets(fstr, data = mtcars, method = "exhaustive", nbest = 2)
# 2. regsubsets.default(fstr, data = mtcars, method = "exhaustive", nbest = 2)
# 3. leaps.setup(x, y, wt = weights, nbest = nbest, nvmax = nvmax, 
#  .     force.in = force.in, force.out = force.out, intercept = intercept)

Answer 1

You're on the right track about the quotes. These indicate that you are passing a character string ( fstr ) rather than a formula to regsubsets . But a character string is not a formula, even if it is a character string of a formula. Turning a character string of a formula into an actual formula is as easy as as.formula(fstr) .

So that minor modification produces this result:

library(leaps)

data(mtcars)

yvar  <- 'mpg'
xvars <-  setdiff(colnames(mtcars), yvar)

fstr <- paste(yvar, '~', paste(setdiff(names(mtcars), yvar), collapse = '+'))
fstr <- as.formula(fstr)

subset = regsubsets(fstr, data = mtcars, method = 'exhaustive', nbest = 2)
res = summary(subset)
res
#> Subset selection object
#> Call: regsubsets.formula(fstr, data = mtcars, method = "exhaustive", 
#>     nbest = 2)
#> 10 Variables  (and intercept)
#>      Forced in Forced out
#> cyl      FALSE      FALSE
#> disp     FALSE      FALSE
#> hp       FALSE      FALSE
#> drat     FALSE      FALSE
#> wt       FALSE      FALSE
#> qsec     FALSE      FALSE
#> vs       FALSE      FALSE
#> am       FALSE      FALSE
#> gear     FALSE      FALSE
#> carb     FALSE      FALSE
#> 2 subsets of each size up to 8
#> Selection Algorithm: exhaustive
#>          cyl disp hp  drat wt  qsec vs  am  gear carb
#> 1  ( 1 ) " " " "  " " " "  "*" " "  " " " " " "  " " 
#> 1  ( 2 ) "*" " "  " " " "  " " " "  " " " " " "  " " 
#> 2  ( 1 ) "*" " "  " " " "  "*" " "  " " " " " "  " " 
#> 2  ( 2 ) " " " "  "*" " "  "*" " "  " " " " " "  " " 
#> 3  ( 1 ) " " " "  " " " "  "*" "*"  " " "*" " "  " " 
#> 3  ( 2 ) "*" " "  "*" " "  "*" " "  " " " " " "  " " 
#> 4  ( 1 ) " " " "  "*" " "  "*" "*"  " " "*" " "  " " 
#> 4  ( 2 ) " " " "  " " " "  "*" "*"  " " "*" " "  "*" 
#> 5  ( 1 ) " " "*"  "*" " "  "*" "*"  " " "*" " "  " " 
#> 5  ( 2 ) " " " "  " " "*"  "*" "*"  " " "*" " "  "*" 
#> 6  ( 1 ) " " "*"  "*" "*"  "*" "*"  " " "*" " "  " " 
#> 6  ( 2 ) " " "*"  "*" " "  "*" "*"  " " "*" "*"  " " 
#> 7  ( 1 ) " " "*"  "*" "*"  "*" "*"  " " "*" "*"  " " 
#> 7  ( 2 ) "*" "*"  "*" "*"  "*" "*"  " " "*" " "  " " 
#> 8  ( 1 ) " " "*"  "*" "*"  "*" "*"  " " "*" "*"  "*" 
#> 8  ( 2 ) " " "*"  "*" "*"  "*" "*"  "*" "*" "*"  " "

^{Created on 2020-12-18 by the reprex package (v0.3.0)}