`standardize()` fails in some cases

Cf https://github.com/easystats/report/issues/379

library(datawizard)

data(sleep)
d <- data_modify(sleep, group = as.integer(group) - 1L)
d_wide <- data_to_wide(
  d,
  names_from = "group",
  values_from = "extra",
  names_prefix = "group"
)

model1 <- lm(d_wide$group0 - d_wide$group1 ~ 1)
standardize(model1)
#> Warning: Using `$` in model formulas can produce unexpected results. Specify your
#>   model using the `data` argument instead.
#>   Try: group0 - group1 ~ 1,
#>   data =

#> Warning: Using `$` in model formulas can produce unexpected results. Specify your
#>   model using the `data` argument instead.
#>   Try: group0 - group1 ~ 1,
#>   data =

#> Warning: Using `$` in model formulas can produce unexpected results. Specify your
#>   model using the `data` argument instead.
#>   Try: group0 - group1 ~ 1,
#>   data =
#> 
#> Call:
#> lm(formula = d_wide$group0 - d_wide$group1 ~ 1, data = data_std)
#> 
#> Coefficients:
#> (Intercept)  
#>       -1.58

model1 <- lm(group0 - group1 ~ 1, data = d_wide)
standardize(model1)
#> Error in eval(predvars, data, env): objet 'group0' introuvable
#> Error: Unable to refit the model with standardized data.
#>   Try instead to standardize the data (standardize(data)) and refit the
#>   model manually.

^{Created on 2023-07-01 with reprex v2.0.2}