Add AUC metrics and improve calibration reporting (#149)

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "FSRS-Optimizer"
-version = "5.4.0"
+version = "5.4.1"
 readme = "README.md"
 dependencies = [
     "matplotlib>=3.7.0",

src/fsrs_optimizer/__main__.py

@@ -146,6 +146,12 @@ def remembered_fallback_prompt(key: str, pretty: str = None):
     print(f"Loss before training: {loss_before:.4f}")
     print(f"Loss after training: {loss_after:.4f}")
     metrics, figures = optimizer.calibration_graph()
+    for partition in metrics:
+        print(f"Last rating = {partition}")
+        for metric in metrics[partition]:
+            print(f"{metric}: {metrics[partition][metric]:.4f}")
+        print()
+
     metrics["Log loss"] = loss_after
     if save_graphs:
         for i, f in enumerate(figures):

src/fsrs_optimizer/fsrs_optimizer.py

@@ -23,6 +23,7 @@
     mean_absolute_error,
     mean_absolute_percentage_error,
     r2_score,
+    roc_auc_score,
 )
 from scipy.optimize import minimize  # type: ignore
 from itertools import accumulate
@@ -1561,10 +1562,17 @@ def calibration_graph(self, dataset=None, verbose=True):
         rmse = rmse_matrix(dataset)
         if verbose:
             tqdm.write(f"RMSE(bins): {rmse:.4f}")
+        metrics_all = {}
         metrics = plot_brier(
             dataset["p"], dataset["y"], bins=20, ax=fig1.add_subplot(111)
         )
-        metrics["rmse"] = rmse
+        metrics["RMSE(bins)"] = rmse
+        metrics["AUC"] = (
+            roc_auc_score(y_true=dataset["y"], y_score=dataset["p"])
+            if len(dataset["y"].unique()) == 2
+            else np.nan
+        )
+        metrics_all["all"] = metrics
         fig2 = plt.figure(figsize=(16, 12))
         for last_rating in (1, 2, 3, 4):
             calibration_data = dataset[dataset["last_rating"] == last_rating]
@@ -1574,13 +1582,23 @@ def calibration_graph(self, dataset=None, verbose=True):
             if verbose:
                 tqdm.write(f"\nLast rating: {last_rating}")
                 tqdm.write(f"RMSE(bins): {rmse:.4f}")
-            plot_brier(
+            metrics = plot_brier(
                 calibration_data["p"],
                 calibration_data["y"],
                 bins=20,
                 ax=fig2.add_subplot(2, 2, int(last_rating)),
                 title=f"Last rating: {last_rating}",
             )
+            metrics["RMSE(bins)"] = rmse
+            metrics["AUC"] = (
+                roc_auc_score(
+                    y_true=calibration_data["y"],
+                    y_score=calibration_data["p"],
+                )
+                if len(calibration_data["y"].unique()) == 2
+                else np.nan
+            )
+            metrics_all[last_rating] = metrics
 
         fig3 = plt.figure()
         self.calibration_helper(
@@ -1611,7 +1629,7 @@ def calibration_graph(self, dataset=None, verbose=True):
             False,
             fig5.add_subplot(111),
         )
-        return metrics, (fig1, fig2, fig3, fig4, fig5)
+        return metrics_all, (fig1, fig2, fig3, fig4, fig5)
 
     def calibration_helper(self, calibration_data, key, bin_func, semilogx, ax1):
         ax2 = ax1.twinx()
@@ -1925,7 +1943,14 @@ def plot_brier(predictions, real, bins=20, ax=None, title=None):
     ax2.legend(loc="lower center")
     if title:
         ax.set_title(title)
-    metrics = {"R-squared": r2, "MAE": mae, "ICI": ici, "E50": e_50, "E90": e_90, "EMax": e_max}
+    metrics = {
+        "R-squared": r2,
+        "MAE": mae,
+        "ICI": ici,
+        "E50": e_50,
+        "E90": e_90,
+        "EMax": e_max,
+    }
     return metrics