setup new datasets study

asreview · Jan 22, 2025 · 0ecd8b2 · 0ecd8b2
1 parent 384b28c
commit 0ecd8b2
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 6 deletions.
diff --git a/asreview2-optuna/classifiers.py b/asreview2-optuna/classifiers.py
@@ -29,9 +29,9 @@ def svm_params(trial: optuna.trial.FrozenTrial):
 
     # Only set gamma to a value if we use rbf kernel
     gamma = "scale"
-    #if kernel == "rbf":
-    #    # Use logarithmic normal distribution for gamma (gamma effect is non-linear)
-    #    gamma = trial.suggest_float("svm__gamma", 1e-4, 10, log=True)
+    if kernel == "rbf":
+        # Use logarithmic normal distribution for gamma (gamma effect is non-linear)
+        gamma = trial.suggest_float("svm__gamma", 1e-4, 10, log=True)
     return {"C": C, "kernel": kernel, "gamma": gamma}
 
 

diff --git a/asreview2-optuna/feature_extractors.py b/asreview2-optuna/feature_extractors.py
@@ -5,7 +5,7 @@
 
 
 def tfidf_params(trial: optuna.trial.FrozenTrial):
-    max_features = trial.suggest_int("tfidf__max_features", 200, 20_000)
+    #max_features = trial.suggest_int("tfidf__max_features", 200, 20_000)
 
     max_df = trial.suggest_float("tfidf__max_df", 0.5, 1.0)
 
@@ -18,7 +18,7 @@ def tfidf_params(trial: optuna.trial.FrozenTrial):
     sublinear_tf = trial.suggest_categorical("tfidf__sublinear_tf", [True, False])
 
     return {
-        "max_features": max_features,
+        #"max_features": max_features,
         "max_df": max_df,
         "min_df": min_df,
         "ngram_range": ngram_range,

diff --git a/asreview2-optuna/main.py b/asreview2-optuna/main.py
@@ -20,7 +20,7 @@
 
 # Study variables
 VERSION = 1
-STUDY_SET = "full"
+STUDY_SET = "demo"
 PICKLE_FOLDER_PATH = Path("synergy-dataset", "pickles")
 CLASSIFIER_TYPE = "svm"  # Options: "nb", "log", "svm", "rf"
 FEATURE_EXTRACTOR_TYPE = "tfidf"  # Options: "tfidf", "onehot"