Use more informative labels

gp_visualizer.py

@@ -34,14 +34,14 @@ class PlotOptions:
     show_training_data: bool = True
     show_true_function: bool = True
     show_mean_prediction: bool = True
-    show_confidence_interval: bool = True
+    show_prediction_interval: bool = True
 
     def update(self, **kwargs):
         return PlotOptions(
             show_training_data=kwargs.get("show_training_data", self.show_training_data),
             show_true_function=kwargs.get("show_true_function", self.show_true_function),
             show_mean_prediction=kwargs.get("show_mean_prediction", self.show_mean_prediction),
-            show_confidence_interval=kwargs.get("show_confidence_interval", self.show_confidence_interval),
+            show_prediction_interval=kwargs.get("show_prediction_interval", self.show_prediction_interval),
         )
 
     def __hash__(self):
@@ -50,7 +50,7 @@ class PlotOptions:
                 self.show_training_data,
                 self.show_true_function,
                 self.show_mean_prediction,
-                self.show_confidence_interval,
+                self.show_prediction_interval,
             )
         )
 
@@ -90,13 +90,13 @@ def eval_kernel(kernel_str) -> Kernel:
 class ModelState:
     model: GaussianProcessRegressor
     kernel: str
-    mode: str
+    distribution: str
 
     def __hash__(self):
         return hash(
             (
                 self.kernel,
-                self.mode,
+                self.distribution,
             )
         )
 
@@ -181,17 +181,17 @@ class GpVisualizer:
                 x_test.flatten(), 
                 y_pred, 
                 linestyle="--", 
-                label='prediction', 
+                label='mean prediction', 
                 color=self.plot_cmap(2),
             )
-        if plot_options.show_confidence_interval and x_test is not None and y_std is not None:
+        if plot_options.show_prediction_interval and x_test is not None and y_std is not None:
             plt.fill_between(
                 x_test.flatten(), 
                 y_pred - 1.96 * y_std, 
                 y_pred + 1.96 * y_std, 
                 color=self.plot_cmap(3), 
                 alpha=0.2,
-                label='95% confidence interval',
+                label='95% prediction interval',
             )
 
         if x_test is not None and sample_y:
@@ -225,14 +225,14 @@ class GpVisualizer:
         self, 
         kernel: str,
         dataset: Dataset,
-        mode: str,
+        distribution: str,
     ) -> GaussianProcessRegressor:
         model = GaussianProcessRegressor(kernel=eval_kernel(kernel))
-        if mode == "posterior":
+        if distribution == "posterior":
             if dataset.x.shape[0] > 0:
                 model.fit(dataset.x, dataset.y)
-        elif mode != "prior":
-            raise ValueError(f"Unknown mode: {mode}")
+        elif distribution != "prior":
+            raise ValueError(f"Unknown distribution: {distribution}")
 
         return model
 
@@ -246,10 +246,10 @@ class GpVisualizer:
         model = self.init_model(
             model_state.kernel,
             dataset,
-            model_state.mode,
+            model_state.distribution,
         )
         model_state = ModelState(
-            model=model, kernel=model_state.kernel, mode=model_state.mode
+            model=model, kernel=model_state.kernel, distribution=model_state.distribution
         )
 
         new_canvas = self.plot(dataset, model_state, plot_options)
@@ -259,7 +259,7 @@ class GpVisualizer:
     def update_model(
         self, 
         kernel_str: str, 
-        mode: str,
+        distribution: str,
         model_state: ModelState,
         dataset: Dataset,
         plot_options: PlotOptions,
@@ -269,10 +269,10 @@ class GpVisualizer:
             model = self.init_model(
                 kernel_str,
                 dataset,
-                mode.lower(),
+                distribution.lower(),
             )
             model_state = ModelState(
-                model=model, kernel=kernel_str, mode=mode.lower()
+                model=model, kernel=kernel_str, distribution=distribution.lower()
             )
         except Exception as e:
             logger.error(f"Error updating kernel: {e}")
@@ -331,7 +331,7 @@ class GpVisualizer:
             kernel = "RBF(length_scale=1.0) + WhiteKernel(noise_level=1.0)"
             model = self.init_model(kernel, dataset.value, "posterior")
             model_state = gr.State(
-                ModelState(model=model, kernel=kernel, mode="posterior")
+                ModelState(model=model, kernel=kernel, distribution="posterior")
             )
 
             # GUI elements and layout 
@@ -358,29 +358,24 @@ class GpVisualizer:
                         )
 
                     with gr.Tab("Model"):
-                        gr.Textbox(
-                            label="Prior Mean", 
-                            value="0",
-                            interactive=False,
-                        )
                         kernel_box = gr.Textbox(
                             label="Kernel", 
                             value=model_state.value.kernel,
                             interactive=True,
                         )
-                        mode = gr.Radio(
-                            label="Mode",
+                        distribution = gr.Radio(
+                            label="Distribution",
                             choices=["Prior", "Posterior"],
                             value="Posterior",
                         )
                         kernel_box.submit(
                             fn=self.update_model,
-                            inputs=[kernel_box, mode, model_state, dataset, plot_options],
+                            inputs=[kernel_box, distribution, model_state, dataset, plot_options],
                             outputs=[model_state, canvas],
                         )
-                        mode.change(
+                        distribution.change(
                             fn=self.update_model,
-                            inputs=[kernel_box, mode, model_state, dataset, plot_options],
+                            inputs=[kernel_box, distribution, model_state, dataset, plot_options],
                             outputs=[model_state, canvas],
                         )
 
@@ -410,7 +405,7 @@ class GpVisualizer:
                             label="Show Mean Prediction", 
                             value=True,
                         )
-                        show_confidence_interval = gr.Checkbox(
+                        show_prediction_interval = gr.Checkbox(
                             label="Show Confidence Interval",
                             value=True,
                         )
@@ -429,9 +424,9 @@ class GpVisualizer:
                             inputs=[show_mean_prediction, plot_options],
                             outputs=[plot_options],
                         )
-                        show_confidence_interval.change(
-                            fn=lambda val, options: options.update(show_confidence_interval=val),
-                            inputs=[show_confidence_interval, plot_options],
+                        show_prediction_interval.change(
+                            fn=lambda val, options: options.update(show_prediction_interval=val),
+                            inputs=[show_prediction_interval, plot_options],
                             outputs=[plot_options],
                         )
                         plot_options.change(