ENH: Add non normalized confusion matrix to plot options (#791)

Add non normalized confusion matrix to plot options

hi-ml-cpath/src/health_cpath/utils/plots_utils.py

@@ -17,7 +17,7 @@ from health_cpath.utils.viz_utils import (
     plot_attention_tiles,
     plot_heatmap_overlay,
     plot_attention_histogram,
-    plot_normalized_confusion_matrix,
+    plot_normalized_and_non_normalized_confusion_matrices,
     plot_scores_hist,
     plot_slide,
 )
@@ -130,15 +130,11 @@ def save_confusion_matrix(results: ResultsType, class_names: Sequence[str], figu
     if pred_labels_diff_expected != set():
         raise ValueError("More entries were found in predicted labels than are available in class names")
 
-    cf_matrix_n = confusion_matrix(
-        true_labels,
-        pred_labels,
-        labels=all_potential_labels,
-        normalize="true"
-    )
+    cf_matrix = confusion_matrix(true_labels, pred_labels, labels=all_potential_labels)
+    cf_matrix_n = confusion_matrix(true_labels, pred_labels, labels=all_potential_labels, normalize="true")
 
-    fig = plot_normalized_confusion_matrix(cm=cf_matrix_n, class_names=(class_names))
-    save_figure(fig=fig, figpath=figures_dir / f"normalized_confusion_matrix_{stage}.png")
+    fig = plot_normalized_and_non_normalized_confusion_matrices(cm=cf_matrix, cm_n=cf_matrix_n, class_names=class_names)
+    save_figure(fig=fig, figpath=figures_dir / f"confusion_matrices_{stage}.png")
 
 
 def save_top_and_bottom_tiles(

hi-ml-cpath/src/health_cpath/utils/viz_utils.py

@@ -30,6 +30,7 @@ from health_cpath.utils.tiles_selection_utils import SlideNode
 def load_image_dict(sample: dict, loading_params: LoadingParams) -> Dict[SlideKey, Any]:
     """
     Load image from metadata dictionary
+
     :param sample: dict describing image metadata. Example:
         {'image_id': ['1ca999adbbc948e69783686e5b5414e4'],
         'image': ['/tmp/datasets/PANDA/train_images/1ca999adbbc948e69783686e5b5414e4.tiff'],
@@ -258,8 +259,9 @@ def plot_attention_histogram(case: str, slide_node: SlideNode, results: Dict[Res
 
 def plot_normalized_confusion_matrix(cm: np.ndarray, class_names: Sequence[str]) -> plt.Figure:
     """Plots a normalized confusion matrix and returns the figure.
-    param cm: Normalized confusion matrix to be plotted.
-    param class_names: List of class names.
+
+    :param cm: Normalized confusion matrix to be plotted.
+    :param class_names: List of class names.
     """
     fig, ax = plt.subplots()
     ax = sns.heatmap(cm, annot=True, cmap="Blues", fmt=".2%")
@@ -270,9 +272,30 @@ def plot_normalized_confusion_matrix(cm: np.ndarray, class_names: Sequence[str])
     return fig
 
 
+def plot_normalized_and_non_normalized_confusion_matrices(
+    cm: np.ndarray, cm_n: np.ndarray, class_names: Sequence[str],
+) -> plt.Figure:
+    """Plots a normalized and non-normalized confusion matrix and returns the figure.
+
+    :param cm: Non normalized confusion matrix to be plotted.
+    :param cm_n: Normalized confusion matrix to be plotted.
+    :param class_names: List of class names.
+    """
+    fig, axes = plt.subplots(1, 2, figsize=(15, 5))
+    axes[0] = sns.heatmap(cm, annot=True, cmap="Blues", fmt="d", ax=axes[0])
+    axes[1] = sns.heatmap(cm_n, annot=True, cmap="Blues", fmt=".2%", ax=axes[1])
+    for ax in axes:
+        ax.set_xlabel("Predicted")
+        ax.set_ylabel("True")
+        ax.xaxis.set_ticklabels(class_names)
+        ax.yaxis.set_ticklabels(class_names)
+    return fig
+
+
 def resize_and_save(width_inch: int, height_inch: int, filename: Union[Path, str], dpi: int = 150) -> None:
     """
     Resizes the present figure to the given (width, height) in inches, and saves it to the given filename.
+
     :param width_inch: The width of the figure in inches.
     :param height_inch: The height of the figure in inches.
     :param filename: The filename to save to.

hi-ml-cpath/testhisto/testhisto/utils/test_plots_utils.py

@@ -131,8 +131,12 @@ def test_save_conf_matrix_integration(tmp_path: Path) -> None:
     class_names = ["foo", "bar"]
 
     save_confusion_matrix(results, class_names, tmp_path, stage='foo')
-    file = Path(tmp_path) / "normalized_confusion_matrix_foo.png"
+    file = Path(tmp_path) / "confusion_matrices_foo.png"
     assert file.exists()
+    expected = full_ml_test_data_path("histo_heatmaps") / "confusion_matrices_foo.png"
+    # To update the stored results, uncomment this line:
+    # expected.write_bytes(file.read_bytes())
+    assert_binary_files_match(file, expected)
 
     # check that an error is raised if true labels include indices greater than the expected number of classes
     invalid_results_1 = {
@@ -156,11 +160,15 @@ def test_save_conf_matrix_integration(tmp_path: Path) -> None:
     class_names_extended = ["foo", "bar", "baz"]
     num_classes = len(class_names_extended)
     expected_conf_matrix_shape = (num_classes, num_classes)
-    with patch("health_cpath.utils.plots_utils.plot_normalized_confusion_matrix") as mock_plot_conf_matrix:
+    with patch(
+        "health_cpath.utils.plots_utils.plot_normalized_and_non_normalized_confusion_matrices"
+    ) as mock_plot_conf_matrix:
         with patch("health_cpath.utils.plots_utils.save_figure"):
             save_confusion_matrix(results, class_names_extended, tmp_path)
             mock_plot_conf_matrix.assert_called_once()
+            actual_n_conf_matrix = mock_plot_conf_matrix.call_args[1].get('cm_n')
             actual_conf_matrix = mock_plot_conf_matrix.call_args[1].get('cm')
+            assert actual_n_conf_matrix.shape == expected_conf_matrix_shape
             assert actual_conf_matrix.shape == expected_conf_matrix_shape