Fix datetime type cast. Fix hinge loss

ThomasMeissnerDS · Sep 12, 2024 · 4c67b51 · 4c67b51
1 parent 8d8c38d
commit 4c67b51
Show file tree

Hide file tree

Showing 3 changed files with 792 additions and 699 deletions.
diff --git a/bluecast/conformal_prediction/nonconformity_measures.py b/bluecast/conformal_prediction/nonconformity_measures.py
@@ -43,7 +43,10 @@ def hinge_loss(
 
     for true_class, preds_arr in zip(y_true, y_hat):
         hinge_losses.append(1 - preds_arr[true_class])
-    return np.asarray(hinge_losses)
+    try:
+        return np.asarray(hinge_losses, dtype="float64")
+    except ValueError:
+        return np.asarray(hinge_losses, dtype="object")
 
 
 def margin_nonconformity_measure(
@@ -71,7 +74,10 @@ def margin_nonconformity_measure(
         probas_y_false = np.delete(preds_arr, true_class)
         prob_y_most_likely_false = np.max(probas_y_false)
         mnm_losses.append(prob_y_true - prob_y_most_likely_false)
-    return np.asarray(mnm_losses)
+    try:
+        return np.asarray(mnm_losses, dtype="float64")
+    except ValueError:
+        return np.asarray(mnm_losses, dtype="object")
 
 
 def brier_score(
@@ -94,4 +100,7 @@ def brier_score(
     brier_losses = []
     for true_class, preds_arr in zip(y_true, y_hat):
         brier_losses.append((1 - preds_arr[true_class]) ** 2)
-    return np.asarray(brier_losses)
+    try:
+        return np.asarray(brier_losses, dtype="float64")
+    except ValueError:
+        return np.asarray(brier_losses, dtype="object")
diff --git a/bluecast/preprocessing/datetime_features.py b/bluecast/preprocessing/datetime_features.py
@@ -32,16 +32,16 @@ def date_converter(
 
     for c in date_columns:
         if "year" in date_parts:
-            df[str(c) + "_year"] = df[c].dt.year.astype(int)
+            df[str(c) + "_year"] = df[c].dt.year.astype(float)
         if "month" in date_parts:
-            df[str(c) + "_month"] = df[c].dt.month.astype(int)
+            df[str(c) + "_month"] = df[c].dt.month.astype(float)
         if "week_of_year" in date_parts:
-            df[str(c) + "_week_of_year"] = df[c].dt.isocalendar().week.astype(int)
+            df[str(c) + "_week_of_year"] = df[c].dt.isocalendar().week.astype(float)
         if "day" in date_parts:
-            df[str(c) + "_day"] = df[c].dt.day.astype(int)
+            df[str(c) + "_day"] = df[c].dt.day.astype(float)
         if "dayofweek" in date_parts:
-            df[str(c) + "_dayofweek"] = df[c].dt.dayofweek.astype(int)
+            df[str(c) + "_dayofweek"] = df[c].dt.dayofweek.astype(float)
         if "hour" in date_parts:
-            df[str(c) + "_hour"] = df[c].dt.hour.astype(int)
+            df[str(c) + "_hour"] = df[c].dt.hour.astype(float)
         df = df.drop(c, axis=1)
     return df