Droplet reconstruction, patch_pixels auto-detect, XAS fixes, and XCS/MFX pipelines

.gitignore

@@ -11,3 +11,6 @@ __pycache__/
 
 # Coverage
 .coverage
+
+# Pipeline output data (HDF5 results, large arrays)
+examples/results/

XSpect/XSpect_Processor/XSpectDetectorProcessor.py

@@ -4,12 +4,13 @@
 import matplotlib.pyplot as plt
 from sklearn.cluster import DBSCAN
 
+
 class XSpectDetectorProcessor:
     """
-    Processor for 2D detector data to perform image normalization, edge detection, 
+    Processor for 2D detector data to perform image normalization, edge detection,
     and alignment of signals.
     """
-    
+
     def __init__(self, image):
         """
         Initialize the processor with the image.
@@ -23,7 +24,7 @@ def __init__(self, image):
         self.image_8bit = self.normalize_to_8bit(image)
         self.edges = None
         self.aligned_image = None
-    
+
     def normalize_to_8bit(self, image):
         """
         Normalize image data to 8-bit range [0, 255].
@@ -46,7 +47,7 @@ def normalize_to_8bit(self, image):
 
         # Convert to 8-bit values
         image_8bit = normalized_image.astype(np.uint8)
-        
+
         return image_8bit
 
     def detect_edges(self, low_threshold=30, high_threshold=100):
@@ -68,7 +69,6 @@ def detect_edges(self, low_threshold=30, high_threshold=100):
         self.edges = cv2.Canny(self.image_8bit, low_threshold, high_threshold)
         return self.edges
 
-
     def find_optimal_rotation_angle(self):
         """
         Calculate the optimal rotation angle to align the signals.
@@ -77,56 +77,55 @@ def find_optimal_rotation_angle(self):
         -------
         float
             Optimal rotation angle in degrees.
-        
+
         Raises
         ------
         ValueError
             If edges have not been detected.
         """
         if self.edges is None:
             raise ValueError("Edges have not been detected. Call `detect_edges` first.")
-        
+
         # Find all non-zero (signal) pixels
         non_zero_coords = np.column_stack(np.nonzero(self.edges))
-        
-        #Identifying multiple XES edges.
+
+        # Identifying multiple XES edges.
         clustering = DBSCAN(eps=5, min_samples=10).fit(non_zero_coords)
         labels = clustering.labels_
-        
+
         unique_labels = set(labels)
         angles = []
-
+        cluster_sizes = []
+
         for label in unique_labels:
             if label == -1:
                 # -1 indicates noise
                 continue
-            
+
             cluster_coords = non_zero_coords[labels == label]
-            
+
             mean = np.mean(cluster_coords, axis=0)
             centered_coords = cluster_coords - mean
             covariance_matrix = np.cov(centered_coords, rowvar=False)
             eigenvalues, eigenvectors = np.linalg.eigh(covariance_matrix)
-
-
+
             principal_vector = eigenvectors[:, np.argmax(eigenvalues)]
-
-
-            angle = np.arctan2(principal_vector[1], principal_vector[0])
-
-
+
+            angle = np.arctan2(principal_vector[0], principal_vector[1])
+
             angle_degrees = np.degrees(angle)
             angles.append(angle_degrees)
+            cluster_sizes.append(len(cluster_coords))
+
+        # Weight by cluster size so small artifacts don't bias the result
+        optimal_angle = float(np.average(angles, weights=cluster_sizes))
+        self.angles = angles
 
-
-        optimal_angle = np.mean(angles)
-        self.angles=angles
-
         if optimal_angle > 90:
             optimal_angle -= 180
         elif optimal_angle < -90:
             optimal_angle += 180
-        
+
         return optimal_angle
 
     def align_image(self):
@@ -141,7 +140,7 @@ def align_image(self):
         optimal_angle = self.find_optimal_rotation_angle()
         self.aligned_image = rotate(self.image, -optimal_angle, reshape=False)
         return self.aligned_image
-    
+
     def plot_images(self):
         """
         Plot the original image, edge-detected image, and aligned image.
@@ -155,22 +154,22 @@ def plot_images(self):
             raise ValueError("Edges have not been detected. Call `detect_edges` first.")
         if self.aligned_image is None:
             raise ValueError("Image has not been aligned. Call `align_image` first.")
-        
-        fig, axes = plt.subplots(1, 3, figsize=(9, 3),dpi=100)
-        
+
+        fig, axes = plt.subplots(1, 3, figsize=(9, 3), dpi=100)
+
         # Original image
-        axes[0].imshow(self.image, cmap='RdBu')
+        axes[0].imshow(self.image, cmap="RdBu")
         axes[0].set_title("Original Image")
-        axes[0].axis('off')
-        
+        axes[0].axis("off")
+
         # Edge-detected image
-        axes[1].imshow(self.edges, cmap='gray')
+        axes[1].imshow(self.edges, cmap="gray")
         axes[1].set_title("XES Detected Image")
-        axes[1].axis('off')
-        
+        axes[1].axis("off")
+
         # Aligned image
-        axes[2].imshow(self.aligned_image, cmap='RdBu')
+        axes[2].imshow(self.aligned_image, cmap="RdBu")
         axes[2].set_title("Aligned Image")
-        axes[2].axis('off')
+        axes[2].axis("off")
         plt.tight_layout()
-        plt.show()
+        plt.show()

XSpect/analysis/__init__.py

@@ -14,3 +14,4 @@
 import XSpect.analysis.spectroscopy  # noqa: F401
 import XSpect.analysis.xes  # noqa: F401
 import XSpect.analysis.xas  # noqa: F401
+import XSpect.analysis.droplet  # noqa: F401