Uninformative error if parsnip extension package isn't loaded

[EmilHvitfeldt]

You get a nice text when you print the model specification.

This isn't a parsnip issue as far as I can tell since you get a nice warning if you try with just a parsnip object.

library(tidymodels)

model_recipe <- recipe(Species ~ ., data = iris)

# Create a workflow
model_final <- naive_Bayes() |>
  set_mode("classification") |>
  set_engine("klaR")

train_fit <- fit(model_final, Species ~ ., data = iris)
#> Error in `fit()`:
#> ! parsnip could not locate an implementation for `naive_Bayes`
#>   classification model specifications using the `klaR` engine.
#> ℹ The parsnip extension package discrim implements support for this
#>   specification.
#> ℹ Please install (if needed) and load to continue.

#> Backtrace:
#>     ▆
#>  1. ├─generics::fit(model_final, Species ~ ., data = iris)
#>  2. └─parsnip::fit.model_spec(model_final, Species ~ ., data = iris)
#>  3.   └─parsnip:::prompt_missing_implementation(...) at parsnip/R/fit.R:121:6
#>  4.     └─cli (local) prompt(c(msg, ""), ...) at parsnip/R/misc.R:201:2
#>  5.       └─rlang::abort(...)

Reprex

library(tidymodels)

model_recipe <- recipe(Species ~ ., data = iris)

# Create a workflow
model_final <- naive_Bayes() |>
  set_mode("classification") |>
  set_engine("klaR")

model_final_wf <- workflow() |>
  add_recipe(model_recipe) |>
  add_model(model_final)

train_fit <- fit(model_final_wf, data = iris)
#> Error in `fit_xy()`:
#> ! data.frame_ is unknown.

#> Backtrace:
#>     ▆
#>  1. ├─generics::fit(model_final_wf, data = iris)
#>  2. └─workflows:::fit.workflow(model_final_wf, data = iris)
#>  3.   └─workflows::.fit_model(workflow, control)
#>  4.     ├─generics::fit(action_model, workflow = workflow, control = control)
#>  5.     └─workflows:::fit.action_model(...)
#>  6.       └─workflows:::fit_from_xy(spec, mold, case_weights, control_parsnip)
#>  7.         ├─generics::fit_xy(...)
#>  8.         └─parsnip::fit_xy.model_spec(...)
#>  9.           └─rlang::abort(glue::glue("{interfaces} is unknown.")) at parsnip/R/fit.R:292:4

model_final_wf
#> ══ Workflow ════════════════════════════════════════════════════════════════════
#> Preprocessor: Recipe
#> Model: naive_Bayes()
#> 
#> ── Preprocessor ────────────────────────────────────────────────────────────────
#> 0 Recipe Steps
#> 
#> ── Model ───────────────────────────────────────────────────────────────────────
#> ! parsnip could not locate an implementation for `naive_Bayes` classification
#>   model specifications using the `klaR` engine.
#> ℹ The parsnip extension package discrim implements support for this
#>   specification.
#> ℹ Please install (if needed) and load to continue.
#> Naive Bayes Model Specification (classification)
#> 
#> Computational engine: klaR

library(discrim)
#> 
#> Attaching package: 'discrim'
#> The following object is masked from 'package:dials':
#> 
#>     smoothness

train_fit <- fit(model_final_wf, data = iris)

train_fit
#> ══ Workflow [trained] ══════════════════════════════════════════════════════════
#> Preprocessor: Recipe
#> Model: naive_Bayes()
#> 
#> ── Preprocessor ────────────────────────────────────────────────────────────────
#> 0 Recipe Steps
#> 
#> ── Model ───────────────────────────────────────────────────────────────────────
#> $apriori
#> grouping
#>     setosa versicolor  virginica 
#>  0.3333333  0.3333333  0.3333333 
#> 
#> $tables
#> $tables$Sepal.Length
#> $tables$Sepal.Length$setosa
#> 
#> Call:
#>  density.default(x = xx)
#> 
#> Data: xx (50 obs.);  Bandwidth 'bw' = 0.1229
#> 
#>        x               y            
#>  Min.   :3.931   Min.   :0.0008217  
#>  1st Qu.:4.491   1st Qu.:0.1008903  
#>  Median :5.050   Median :0.3796642  
#>  Mean   :5.050   Mean   :0.4465351  
#>  3rd Qu.:5.609   3rd Qu.:0.6865548  
#>  Max.   :6.169   Max.   :1.2397687  
#> 
#> $tables$Sepal.Length$versicolor
#> 
#> Call:
#>  density.default(x = xx)
#> 
#> Data: xx (50 obs.);  Bandwidth 'bw' = 0.2124
#> 
#>        x               y            
#>  Min.   :4.263   Min.   :0.0005386  
#>  1st Qu.:5.106   1st Qu.:0.0508068  
#>  Median :5.950   Median :0.2503411  
#>  Mean   :5.950   Mean   :0.2960165  
#>  3rd Qu.:6.794   3rd Qu.:0.5117124  
#>  Max.   :7.637   Max.   :0.7319843  
#> 
#> $tables$Sepal.Length$virginica
#> 
#> Call:
#>  density.default(x = xx)
#> 
#> Data: xx (50 obs.);  Bandwidth 'bw' = 0.2073
#> 
#>        x               y            
#>  Min.   :4.278   Min.   :0.0004334  
#>  1st Qu.:5.339   1st Qu.:0.0304136  
#>  Median :6.400   Median :0.2147562  
#>  Mean   :6.400   Mean   :0.2353874  
#>  3rd Qu.:7.461   3rd Qu.:0.3752514  
#> 
#> ...
#> and 312 more lines.

^{Created on 2022-10-25 with reprex v2.0.2}

[simonpcouch]

Ah, thanks! Duplicate of #174, perhaps we ought to reconsider #175.

[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.