Merge pull request #482 from NatLibFi/fix-mllm-hyperopt-hidden-labels

MLLM fix: don't include use_hidden_labels in hyperopt, it won't have any effect

annif/backend/mllm.py

@@ -32,8 +32,6 @@ def _objective(self, trial):
             'min_samples_leaf': trial.suggest_int('min_samples_leaf', 5, 30),
             'max_leaf_nodes': trial.suggest_int('max_leaf_nodes', 100, 2000),
             'max_samples': trial.suggest_float('max_samples', 0.5, 1.0),
-            'use_hidden_labels':
-                trial.suggest_categorical('use_hidden_labels', [True, False]),
             'limit': 100
         }
         model = self._backend._model._create_classifier(params)
@@ -59,8 +57,7 @@ def _postprocess(self, study):
         lines = [
             f"min_samples_leaf={bp['min_samples_leaf']}",
             f"max_leaf_nodes={bp['max_leaf_nodes']}",
-            f"max_samples={bp['max_samples']:.4f}",
-            f"use_hidden_labels={bp['use_hidden_labels']}"
+            f"max_samples={bp['max_samples']:.4f}"
         ]
         return hyperopt.HPRecommendation(lines=lines, score=study.best_value)