Bugfix so multilabel confusion matrix can plot for 2 or more labels (#…

…2858)

* added matplotlib to test requirements
* added new test for plotting in multilabel classifier
* added bugfix
* fix errors
* changelog

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Nicki Skafte Detlefsen <skaftenicki@gmail.com>
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

CHANGELOG.md

@@ -27,6 +27,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Fixed
 
+- Fixed plotting of multilabel confusion matrix ([#2858](https://github.com/PyTorchLightning/metrics/pull/2858))
+
+
 - Delete `Device2Host` caused by comm with device and host ([#2840](https://github.com/PyTorchLightning/metrics/pull/2840))
 
 

src/torchmetrics/utilities/plot.py

@@ -270,7 +270,7 @@ def plot_confusion_matrix(
     fig, axs = plt.subplots(nrows=rows, ncols=cols, constrained_layout=True) if ax is None else (ax.get_figure(), ax)
     axs = trim_axs(axs, nb)
     for i in range(nb):
-        ax = axs[i] if rows != 1 and cols != 1 else axs
+        ax = axs[i] if (rows != 1 or cols != 1) else axs
         if fig_label is not None:
             ax.set_title(f"Label {fig_label[i]}", fontsize=15)
         im = ax.imshow(confmat[i].cpu().detach() if confmat.ndim == 3 else confmat.cpu().detach(), cmap=cmap)

tests/unittests/classification/test_confusion_matrix.py

@@ -393,6 +393,16 @@ def test_multilabel_confusion_matrix_dtype_gpu(self, inputs, dtype):
             dtype=dtype,
         )
 
+    @pytest.mark.parametrize("num_labels", [2, NUM_CLASSES])
+    def test_multilabel_confusion_matrix_plot(self, num_labels, inputs):
+        """Test multilabel cm plots."""
+        multi_label_confusion_matrix = MultilabelConfusionMatrix(num_labels=num_labels)
+        preds = target = torch.ones(1, num_labels).int()
+        multi_label_confusion_matrix.update(preds, target)
+        fig, ax = multi_label_confusion_matrix.plot()
+        assert fig is not None
+        assert ax is not None
+
 
 def test_warning_on_nan():
     """Test that a warning is given if division by zero happens during normalization of confusion matrix."""