signal-grad-cam 2.0.0__tar.gz → 2.0.2__tar.gz

{signal_grad_cam-2.0.0/signal_grad_cam.egg-info → signal_grad_cam-2.0.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: signal_grad_cam
-Version: 2.0.0
+Version: 2.0.2
 Summary: SignalGrad-CAM aims at generalising Grad-CAM to one-dimensional applications, while enhancing usability and efficiency.
 Home-page: https://github.com/samuelepe11/signal_grad_cam
 Author: Samuele Pe
@@ -159,7 +159,7 @@ See the [open issues](https://github.com/bmi-labmedinfo/signal_grad_cam/issues)
 ## <i>NEW!</i> Updates in SignalGrad-CAM
 <p align="justify">Compared to previous versions, SignalGrad-CAM now offers the following enhancements:</p>
 
-* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models.</p>
+* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models, now it is only required to set as True the parameter <i>`is_regression_network`</i> in the constructor function.</p>
 * <p align="justify"><i>Contrastive explanations:</i> Users can generate and visualize contrastive explanations by specifying one or more foil classes via the parameter <i>`contrastive_foil_classes`</i>.</p>
 * <p align="justify"><i>3D-CNN support for videos and volumetric data:</i> After expliciting the time axis in the constructor with the parameter <i>`time_axs`</i>, the same functions used for 1D and 2D data now work seamlessly for 3D-CNNs. Outputs include GIF files for quick visualization of 3D activation maps. For a more detailed analysis, users can also request separate PNG images for each volume slice (across the indicated time axis) or video frame using the parameter <i>`show_single_video_frames`</i>.</p>
 

{signal_grad_cam-2.0.0 → signal_grad_cam-2.0.2}/README.md

@@ -136,7 +136,7 @@ See the [open issues](https://github.com/bmi-labmedinfo/signal_grad_cam/issues)
 ## <i>NEW!</i> Updates in SignalGrad-CAM
 <p align="justify">Compared to previous versions, SignalGrad-CAM now offers the following enhancements:</p>
 
-* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models.</p>
+* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models, now it is only required to set as True the parameter <i>`is_regression_network`</i> in the constructor function.</p>
 * <p align="justify"><i>Contrastive explanations:</i> Users can generate and visualize contrastive explanations by specifying one or more foil classes via the parameter <i>`contrastive_foil_classes`</i>.</p>
 * <p align="justify"><i>3D-CNN support for videos and volumetric data:</i> After expliciting the time axis in the constructor with the parameter <i>`time_axs`</i>, the same functions used for 1D and 2D data now work seamlessly for 3D-CNNs. Outputs include GIF files for quick visualization of 3D activation maps. For a more detailed analysis, users can also request separate PNG images for each volume slice (across the indicated time axis) or video frame using the parameter <i>`show_single_video_frames`</i>.</p>
 

{signal_grad_cam-2.0.0 → signal_grad_cam-2.0.2}/setup.py

@@ -5,7 +5,7 @@ with open("README.md", "r") as f:
 
 setup(
     name="signal_grad_cam",
-    version="2.0.0",
+    version="2.0.2",
     description="SignalGrad-CAM aims at generalising Grad-CAM to one-dimensional applications, while enhancing usability"
                 " and efficiency.",
     keywords="XAI, class activation maps, CNN, time series",

{signal_grad_cam-2.0.0 → signal_grad_cam-2.0.2}/signal_grad_cam/cam_builder.py

@@ -108,7 +108,7 @@ class CamBuilder:
                              "overconfident predictions can cause the Sigmoid to saturate, leading to empty or null "
                              "maps. To prevent this, modify your network to output logits directly.\n"
                              "Make sure the provided model is set to inference mode ('eval') if using PyTorch. For "
-                             "TensorFlow/Keras models, ensure the model is built—i.e., it must have defined 'inputs' "
+                             "TensorFlow/Keras models, ensure the model is built — i.e., it must have defined 'inputs' "
                              "and 'output' attributes.\n"
                              "Network layers found (name: type):")
         self._get_layers_pool(show=True, extend_search=extend_search)
@@ -732,9 +732,9 @@ class CamBuilder:
                     w_idx = 1
                     h_idx = 2
                 if not self.input_transposed:
-                    dim_reshape = (data_shape_list[i][0], data_shape_list[i][w_idx], data_shape_list[i][h_idx])
+                    dim_reshape = (data_shape_list[i][self.time_axs], data_shape_list[i][w_idx], data_shape_list[i][h_idx])
                 else:
-                    dim_reshape = (data_shape_list[i][0], data_shape_list[i][h_idx], data_shape_list[i][w_idx])
+                    dim_reshape = (data_shape_list[i][self.time_axs], data_shape_list[i][h_idx], data_shape_list[i][w_idx])
             else:
                 dim_reshape = (1, data_shape_list[i][self.time_axs])
                 if self.time_axs:
@@ -1318,7 +1318,7 @@ class CamBuilder:
         return norm
 
     @staticmethod
-    def __print_justify(text: str, n_characters: int = 170) -> None:
+    def __print_justify(text: str, n_characters: int = 100) -> None:
         """
         Prints a message in a fully justified format within a specified line width.
 
@@ -1345,15 +1345,14 @@ class CamBuilder:
 
         :return:
             - frames: A list of np.ndarrays representing frames to be saved as a GIF animation. This parameter should be
-            provided only for 3D (video/volume) explanations and it is set to None for 1D or 2D
-                maps.
+                provided only for 3D (video/volume) explanations and it is set to None for 1D or 2D maps.
         """
         frames = []
         norm = plt.Normalize(vmin=cam.min(), vmax=cam.max())
         time_axs = 0 if self.time_axs is None else self.time_axs
         for idx in range(cam.shape[time_axs]):
             video_slice = np.take(video, idx, axis=time_axs) if video is not None else None
-            cam_slice = np.take(cam, idx, axis=time_axs)
+            cam_slice = cam[0]
             cam_slice = (plt.get_cmap("inferno")(cam_slice / 255)[:, :, :3] * 255).astype(np.uint8)
             fig, ax = plt.subplots(figsize=(10, 4), dpi=300)
             ax.axis("off")

{signal_grad_cam-2.0.0 → signal_grad_cam-2.0.2/signal_grad_cam.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: signal_grad_cam
-Version: 2.0.0
+Version: 2.0.2
 Summary: SignalGrad-CAM aims at generalising Grad-CAM to one-dimensional applications, while enhancing usability and efficiency.
 Home-page: https://github.com/samuelepe11/signal_grad_cam
 Author: Samuele Pe
@@ -159,7 +159,7 @@ See the [open issues](https://github.com/bmi-labmedinfo/signal_grad_cam/issues)
 ## <i>NEW!</i> Updates in SignalGrad-CAM
 <p align="justify">Compared to previous versions, SignalGrad-CAM now offers the following enhancements:</p>
 
-* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models.</p>
+* <p align="justify"><i>Support for regression tasks:</i> SGrad-CAM can now handle regression-based models. Previously, substantial adjustments were required for these tasks, similar to those still needed for segmentation or generative models, now it is only required to set as True the parameter <i>`is_regression_network`</i> in the constructor function.</p>
 * <p align="justify"><i>Contrastive explanations:</i> Users can generate and visualize contrastive explanations by specifying one or more foil classes via the parameter <i>`contrastive_foil_classes`</i>.</p>
 * <p align="justify"><i>3D-CNN support for videos and volumetric data:</i> After expliciting the time axis in the constructor with the parameter <i>`time_axs`</i>, the same functions used for 1D and 2D data now work seamlessly for 3D-CNNs. Outputs include GIF files for quick visualization of 3D activation maps. For a more detailed analysis, users can also request separate PNG images for each volume slice (across the indicated time axis) or video frame using the parameter <i>`show_single_video_frames`</i>.</p>