## # A tibble: 4 × 4
## spl enrollment_size n percent
## <fct> <chr> <int> <dbl>
## 1 Exceeding or Meeting Expectations large 200 24.4
## 2 Exceeding or Meeting Expectations small 255 31.1
## 3 Partially Meeting or Not Meeting Expectations large 211 25.7
## 4 Partially Meeting or Not Meeting Expectations small 155 18.9
## 25.866 sec elapsed
The initial model run generates an roc_auc of 0.678
## # A tibble: 1 × 6
## .metric .estimator mean n std_err .config
## <chr> <chr> <dbl> <int> <dbl> <chr>
## 1 roc_auc binary 0.678 10 0.0213 Preprocessor1_Model1
## Random Forest Model Specification (classification)
##
## Main Arguments:
## mtry = tune()
## trees = 10000
## min_n = tune()
##
## Engine-Specific Arguments:
## num.threads = cores
## importance = permutation
## verbose = TRUE
##
## Computational engine: ranger
##
## Model fit template:
## ranger::ranger(x = missing_arg(), y = missing_arg(), weights = missing_arg(),
## mtry = min_cols(~tune(), x), num.trees = 10000, min.node.size = min_rows(~tune(),
## x), num.threads = cores, importance = "permutation",
## verbose = TRUE, seed = sample.int(10^5, 1), probability = TRUE)
## 333.297 sec elapsed
With tuning, the roc_auc value improves to 0.683
## # A tibble: 5 × 8
## mtry min_n .metric .estimator mean n std_err .config
## <int> <int> <chr> <chr> <dbl> <int> <dbl> <chr>
## 1 5 40 roc_auc binary 0.680 9 0.0242 Preprocessor1_Model05
## 2 6 20 roc_auc binary 0.679 9 0.0238 Preprocessor1_Model10
## 3 12 30 roc_auc binary 0.679 9 0.0245 Preprocessor1_Model01
## 4 17 33 roc_auc binary 0.675 9 0.0248 Preprocessor1_Model02
## 5 9 13 roc_auc binary 0.673 9 0.0230 Preprocessor1_Model06
## ══ Workflow ════════════════════════════════════════════════════════════════════
## Preprocessor: Recipe
## Model: rand_forest()
##
## ── Preprocessor ────────────────────────────────────────────────────────────────
## 5 Recipe Steps
##
## • step_novel()
## • step_unknown()
## • step_impute_median()
## • step_dummy()
## • step_nzv()
##
## ── Model ───────────────────────────────────────────────────────────────────────
## Random Forest Model Specification (classification)
##
## Main Arguments:
## mtry = 5
## trees = 10000
## min_n = 40
##
## Engine-Specific Arguments:
## num.threads = cores
## importance = permutation
## verbose = TRUE
##
## Computational engine: ranger
## 7.026 sec elapsed
With the workflow, the roc_auc generates a value of 0.626
## [[1]]
## # A tibble: 2 × 4
## .metric .estimator .estimate .config
## <chr> <chr> <dbl> <chr>
## 1 accuracy binary 0.621 Preprocessor1_Model1
## 2 roc_auc binary 0.621 Preprocessor1_Model1
## # A tibble: 206 × 1
## .pred_class
## <fct>
## 1 Exceeding or Meeting Expectations
## 2 Partially Meeting or Not Meeting Expectations
## 3 Exceeding or Meeting Expectations
## 4 Partially Meeting or Not Meeting Expectations
## 5 Partially Meeting or Not Meeting Expectations
## 6 Exceeding or Meeting Expectations
## 7 Partially Meeting or Not Meeting Expectations
## 8 Partially Meeting or Not Meeting Expectations
## 9 Exceeding or Meeting Expectations
## 10 Exceeding or Meeting Expectations
## # … with 196 more rows
## 6.925 sec elapsed
The initial model run generates an roc_auc of 0.594
## # A tibble: 1 × 6
## .metric .estimator mean n std_err .config
## <chr> <chr> <dbl> <int> <dbl> <chr>
## 1 roc_auc binary 0.594 10 0.0157 Preprocessor1_Model1
## Random Forest Model Specification (classification)
##
## Main Arguments:
## mtry = tune()
## trees = 10000
## min_n = tune()
##
## Engine-Specific Arguments:
## num.threads = cores
## importance = permutation
## verbose = TRUE
##
## Computational engine: ranger
##
## Model fit template:
## ranger::ranger(x = missing_arg(), y = missing_arg(), weights = missing_arg(),
## mtry = min_cols(~tune(), x), num.trees = 10000, min.node.size = min_rows(~tune(),
## x), num.threads = cores, importance = "permutation",
## verbose = TRUE, seed = sample.int(10^5, 1), probability = TRUE)
## 219.46 sec elapsed
With tuning, the roc_auc value improves to 0.611
## # A tibble: 5 × 8
## mtry min_n .metric .estimator mean n std_err .config
## <int> <int> <chr> <chr> <dbl> <int> <dbl> <chr>
## 1 1 6 roc_auc binary 0.611 10 0.0165 Preprocessor1_Model07
## 2 3 40 roc_auc binary 0.595 10 0.0160 Preprocessor1_Model05
## 3 4 20 roc_auc binary 0.584 10 0.0131 Preprocessor1_Model10
## 4 5 13 roc_auc binary 0.580 10 0.0120 Preprocessor1_Model06
## 5 7 30 roc_auc binary 0.574 10 0.0121 Preprocessor1_Model01
## ══ Workflow ════════════════════════════════════════════════════════════════════
## Preprocessor: Recipe
## Model: rand_forest()
##
## ── Preprocessor ────────────────────────────────────────────────────────────────
## 5 Recipe Steps
##
## • step_novel()
## • step_unknown()
## • step_impute_median()
## • step_dummy()
## • step_nzv()
##
## ── Model ───────────────────────────────────────────────────────────────────────
## Random Forest Model Specification (classification)
##
## Main Arguments:
## mtry = 1
## trees = 10000
## min_n = 6
##
## Engine-Specific Arguments:
## num.threads = cores
## importance = permutation
## verbose = TRUE
##
## Computational engine: ranger
## 2.266 sec elapsed
With the workflow, the roc_auc generates a value of 0.621
## [[1]]
## # A tibble: 2 × 4
## .metric .estimator .estimate .config
## <chr> <chr> <dbl> <chr>
## 1 accuracy binary 0.578 Preprocessor1_Model1
## 2 roc_auc binary 0.553 Preprocessor1_Model1
## # A tibble: 206 × 1
## .pred_class
## <fct>
## 1 Exceeding or Meeting Expectations
## 2 Partially Meeting or Not Meeting Expectations
## 3 Exceeding or Meeting Expectations
## 4 Partially Meeting or Not Meeting Expectations
## 5 Partially Meeting or Not Meeting Expectations
## 6 Exceeding or Meeting Expectations
## 7 Partially Meeting or Not Meeting Expectations
## 8 Partially Meeting or Not Meeting Expectations
## 9 Exceeding or Meeting Expectations
## 10 Exceeding or Meeting Expectations
## # … with 196 more rows