exported filtered dataset

coralnet_toolbox/MachineLearning/ExportDataset/QtBase.py

@@ -46,7 +46,7 @@ def __init__(self, main_window, parent=None):
         self.selected_labels = []
         self.selected_annotations = []
         self.updating_summary_statistics = False
-        
+
         self.output_dir = None
         self.dataset_name = None
         self.train_ratio = 0.7
@@ -77,7 +77,7 @@ def showEvent(self, event):
         self.update_annotation_type_checkboxes()
         self.populate_class_filter_list()
         self.update_summary_statistics()
-        
+
     def setup_info_layout(self):
         """
         Set up the layout and widgets for the info layout.
@@ -300,7 +300,7 @@ def get_class_mapping(self):
         """
         # Get the label objects for the selected labels
         class_mapping = {}
-        
+
         for label in self.main_window.label_window.labels:
             if label.short_label_code in self.selected_labels:
                 class_mapping[label.short_label_code] = label.to_dict()
@@ -364,7 +364,7 @@ def filter_annotations(self):
         # Filter annotations based on the selected image option
         if self.filtered_images_radio.isChecked():
             annotations = [a for a in annotations if a.image_path in self.image_window.filtered_image_paths]
-        
+
         return annotations
 
     def populate_class_filter_list(self):
@@ -529,6 +529,13 @@ def check_label_distribution(self):
 
         return True
 
+    def update_image_selection(self):
+        """
+        Update the table based on the selected image option.
+        """
+        self.selected_annotations = self.filter_annotations()
+        self.update_summary_statistics()
+
     def update_summary_statistics(self):
         """
         Update the summary statistics for the dataset creation.
@@ -592,7 +599,7 @@ def update_summary_statistics(self):
         self.ready_label.setText("✅ Ready" if (self.ready_status and self.split_status) else "❌ Not Ready")
 
         self.updating_summary_statistics = False
-        
+
     def is_ready(self):
         """Check if the dataset is ready to be created."""
         # Extract the input values, store them in the class variables
@@ -620,9 +627,9 @@ def is_ready(self):
                                 "Input Error",
                                 "Train, Validation, and Test ratios must sum to 1.0")
             return False
-        
+
         return True
-    
+
     def accept(self):
         """
         Handle the OK button click event to create the dataset.
@@ -632,7 +639,7 @@ def accept(self):
 
         # Create the output folder
         output_dir_path = os.path.join(self.output_dir, self.dataset_name)
-        
+
         # Check if the output directory exists
         if os.path.exists(output_dir_path):
             reply = QMessageBox.question(self,
@@ -670,16 +677,9 @@ def accept(self):
                                 "Dataset Created",
                                 "Dataset has been successfully created.")
         super().accept()
-
-    def create_dataset(self):
-        raise NotImplementedError("Method must be implemented in the subclass.")
-
-    def process_annotations(self):
+
+    def create_dataset(self, output_dir_path):
         raise NotImplementedError("Method must be implemented in the subclass.")
 
-    def update_image_selection(self):
-        """
-        Update the table based on the selected image option.
-        """
-        self.selected_annotations = self.filter_annotations()
-        self.update_summary_statistics()
+    def process_annotations(self, annotations, split_dir, split):
+        raise NotImplementedError("Method must be implemented in the subclass.")

coralnet_toolbox/QtAnnotationWindow.py

@@ -408,16 +408,15 @@ def select_annotation(self, annotation, ctrl_pressed=False):
         if len(self.selected_annotations) > 1:
             self.main_window.label_window.deselect_active_label()
             self.main_window.confidence_window.clear_display()
-
-    # TODO test
+
     def select_annotations(self):
         """Select all annotations in the current image."""
         annotations = self.get_image_annotations()
         for annotation in annotations:
             if self.main_window.label_window.label_locked:
                 if annotation.label.id == self.main_window.label_window.locked_label.id:
                     self.select_annotation(annotation, ctrl_pressed=True)
-            else:       
+            else:
                 self.select_annotation(annotation, ctrl_pressed=True)
 
     def unselect_annotation(self, annotation):

coralnet_toolbox/QtEventFilter.py

@@ -75,13 +75,13 @@ def eventFilter(self, obj, event):
                     self.image_window.cycle_next_image()
                     return True
 
-            # Select all annotations on < key press
-            if event.key() == Qt.Key_Less:
+            # Select all annotations on < key press with Shift+Ctrl
+            if event.key() == Qt.Key_Less and event.modifiers() == (Qt.ShiftModifier | Qt.ControlModifier):
                 self.annotation_window.select_annotations()
                 return True
 
-            # Unselect all annotations on > key press
-            if event.key() == Qt.Key_Greater:
+            # Unselect all annotations on > key press with Shift+Ctrl
+            if event.key() == Qt.Key_Greater and event.modifiers() == (Qt.ShiftModifier | Qt.ControlModifier):
                 self.annotation_window.unselect_annotations()
                 return True
 

docs/usage.md

@@ -135,8 +135,8 @@ The main window consists of several components:
 - **Ctrl + Mouse Wheel**: Adjust annotation size.
 - **Ctrl + Left/Right**: Cycle through annotations.
 - **Ctrl + Up/Down**: Cycle through images.
-- **Ctrl + <**: Select all annotations.
-- **Ctrl + >**: Unselect all annotations.
+- **Ctrl + Shift + <**: Select all annotations.
+- **Ctrl + Shift + >**: Unselect all annotations.
 - **Escape**: Exit the program.
 
 - **Machine Learning, SAM, and AutoDistill**: After a model is loaded