Harmonize error for wrong prediction type

[juliasilge]

Closes #480

This PR adds a new little helper function check_spec_pred_type() to create a consistent error when a user uses an unavailable type at prediction time.

library(parsnip)
data(two_class_dat, package = "modeldata")

svm_fit <- svm_linear() %>%
  set_mode("classification") %>%
  set_engine("LiblineaR") %>%
  fit(Class ~ ., data = two_class_dat)

glm_fit <- logistic_reg() %>%
  set_engine("glm") %>%
  fit(Class ~ ., data = two_class_dat)

predict(svm_fit, two_class_dat, type = "prob")
#> Error: No prob prediction method available for this model.
#> * Value for `type` should be one of: 'class', 'raw'
predict(glm_fit, two_class_dat, type = "quantile")
#> Error: No quantile prediction method available for this model.
#> * Value for `type` should be one of: 'class', 'prob', 'raw', 'conf_int'

^{Created on 2021-05-12 by the reprex package (v2.0.0)}

This PR also removes the unneeded class probability prediction module for the LiblineaR svm_linear() which only was there to error.

[github-actions]

This pull request 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.