can't `predict()` with xgboost if fit with `fit_xy()` if data contains outcome

[EmilHvitfeldt]

From some digging. prepare_data() doesn't remove outcomes y if model was fit with fit_xy(). This because an issue here because {xgboost} doesn't like that, while other models doesn't care.

library(parsnip)

spec <- boost_tree() %>%
  set_mode("regression") %>%
  set_engine("xgboost")

lm_fit <- fit(spec, mpg ~ ., data = mtcars)

predict(lm_fit, mtcars)
#> # A tibble: 32 × 1
#>    .pred
#>    <dbl>
#>  1  20.9
#>  2  20.9
#>  3  22.6
#>  4  21.0
#>  5  18.4
#>  6  18.1
#>  7  14.2
#>  8  23.7
#>  9  22.4
#> 10  18.9
#> # ℹ 22 more rows

lm_fit <- fit_xy(spec, x = mtcars[, -1], y = mtcars[, 1])

predict(lm_fit, mtcars)
#> Error in predict.xgb.Booster(object, new_data, ...): Feature names stored in `object` and `newdata` are different!

^{Created on 2024-08-30 with reprex v2.1.0}

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.