nettracer3d 0.8.3__py3-none-any.whl → 0.8.5__py3-none-any.whl

nettracer3d/nettracer.py

@@ -31,6 +31,7 @@ from . import community_extractor
 from . import network_analysis
 from . import morphology
 from . import proximity
+from skimage.segmentation import watershed as water
 
 
 #These next several methods relate to searching with 3D objects by dilating each one in a subarray around their neighborhood although I don't explicitly use this anywhere... can call them deprecated although I may want to use them later again so I have them still written out here.
@@ -104,7 +105,7 @@ def process_label(args):
     print(f"Processing node {label}")
     
     # Get the pre-computed bounding box for this label
-    slice_obj = bounding_boxes[label-1]  # -1 because label numbers start at 1
+    slice_obj = bounding_boxes[int(label)-1]  # -1 because label numbers start at 1
     if slice_obj is None:
         return None, None, None
         
@@ -129,7 +130,7 @@ def create_node_dictionary(nodes, edges, num_nodes, dilate_xy, dilate_z):
     with ThreadPoolExecutor(max_workers=mp.cpu_count()) as executor:
         # Create args list with bounding_boxes included
         args_list = [(nodes, edges, i, dilate_xy, dilate_z, array_shape, bounding_boxes) 
-                    for i in range(1, num_nodes + 1)]
+                    for i in range(1, int(num_nodes) + 1)]
 
         # Execute parallel tasks to process labels
         results = executor.map(process_label, args_list)
@@ -214,7 +215,6 @@ def establish_connections_parallel(edge_labels, num_edge, node_labels):
 
             edge_connections.append(node_labels[index])
 
-        #the set() wrapper removes duplicates from the same sublist
         my_connections = list(set(edge_connections))
 
 
@@ -282,8 +282,36 @@ def extract_pairwise_connections(connections):
 
 
 #Saving outputs
+def create_and_save_dataframe(pairwise_connections, excel_filename=None):
+    """Internal method used to convert lists of discrete connections into an excel output"""
+    
+    # Create DataFrame directly from the connections with 3 columns
+    df = pd.DataFrame(pairwise_connections, columns=['Node A', 'Node B', 'Edge C'])
+    
+    if excel_filename is not None:
+        # Remove file extension if present to use as base path
+        base_path = excel_filename.rsplit('.', 1)[0]
+        
+        # First try to save as CSV
+        try:
+            csv_path = f"{base_path}.csv"
+            df.to_csv(csv_path, index=False)
+            print(f"Network file saved to {csv_path}")
+            return
+        except Exception as e:
+            print(f"Could not save as CSV: {str(e)}")
+            
+            # If CSV fails, try to save as Excel
+            try:
+                xlsx_path = f"{base_path}.xlsx"
+                df.to_excel(xlsx_path, index=False)
+                print(f"Network file saved to {xlsx_path}")
+            except Exception as e:
+                print(f"Unable to write network file to disk... please make sure that {base_path}.xlsx is being saved to a valid directory and try again")
+    else:
+        return df
 
-def create_and_save_dataframe(pairwise_connections, excel_filename = None):
+def create_and_save_dataframe_old(pairwise_connections, excel_filename = None):
     """Internal method used to convert lists of discrete connections into an excel output"""
     # Determine the length of the input list
     length = len(pairwise_connections)
@@ -636,28 +664,25 @@ def threshold(arr, proportion, custom_rad = None):
     def find_closest_index(target: float, num_list: list[float]) -> int:
        return min(range(len(num_list)), key=lambda i: abs(num_list[i] - target))
 
-    # Step 1: Flatten the array
-    flattened = arr.flatten()
 
-    # Step 2: Filter out the zero values
-    non_zero_values = list(set(flattened[flattened > 0]))
+    if custom_rad is not None:
+
+        threshold_value = custom_rad
 
-    # Step 3: Sort the remaining values
-    sorted_values = np.sort(non_zero_values)
+    else:
+        # Step 1: Flatten the array
+        flattened = arr.flatten()
 
-    # Step 4: Determine the threshold for the top proportion%
+        # Step 2: Filter out the zero values
+        non_zero_values = list(set(flattened[flattened > 0]))
 
-    if custom_rad is None:
+        # Step 3: Sort the remaining values
+        sorted_values = np.sort(non_zero_values)
 
         threshold_index = int(len(sorted_values) * proportion)
         threshold_value = sorted_values[threshold_index]
+        print(f"Thresholding as if smallest_radius as assigned {threshold_value}")
 
-    else:
-
-        targ = int(find_closest_index(custom_rad, sorted_values) - (0.02 * len(sorted_values)))
-
-        threshold_value = sorted_values[targ]
-        print(f"Suggested proportion for rad {custom_rad} -> {targ/len(sorted_values)}")
 
     mask = arr > threshold_value
 
@@ -1574,6 +1599,187 @@ def directory_info(directory = None):
     return items
 
 
+# Ripley's K Helpers:
+
+def mirror_points_for_edge_correction(points_array, bounds, max_r, dim=3):
+    """
+    Mirror points near boundaries to handle edge effects in Ripley's K analysis.
+    Works with actual coordinate positions, not spatial grid placement.
+    
+    Parameters:
+    points_array: numpy array of shape (n, 3) with [z, y, x] coordinates (already scaled)
+    bounds: tuple of (min_coords, max_coords) where each is array - can be 2D or 3D
+    max_r: maximum search radius (determines mirroring distance)
+    dim: dimension (2 or 3) - affects which coordinates are used
+    
+    Returns:
+    numpy array with original points plus mirrored points
+    """
+    min_coords, max_coords = bounds
+    
+    # Ensure bounds are numpy arrays and handle dimension mismatch
+    min_coords = np.array(min_coords)
+    max_coords = np.array(max_coords)
+    
+    # Handle case where bounds might be 2D but points are 3D
+    if len(min_coords) == 2 and points_array.shape[1] == 3:
+        # Extend 2D bounds to 3D by adding z=0 dimension at the front
+        min_coords = np.array([0, min_coords[0], min_coords[1]])  # [0, min_x, min_y] -> [min_z, min_y, min_x]
+        max_coords = np.array([0, max_coords[0], max_coords[1]])  # [0, max_x, max_y] -> [max_z, max_y, max_x]
+    elif len(min_coords) == 3 and points_array.shape[1] == 3:
+        # Already 3D, but ensure it's in [z,y,x] format (your bounds are [x,y,z] and get flipped)
+        pass  # Should already be handled by the flip in your bounds calculation
+    
+    # Start with original points
+    all_points = points_array.copy()
+    
+    if dim == 2:
+        # For 2D: work with y, x coordinates (indices 1, 2), z should be 0
+        active_dims = [1, 2]  # y, x
+        # 8 potential mirror regions for 2D (excluding center)
+        mirror_combinations = [
+            [0, -1], [0, 1],   # left, right (y direction)
+            [-1, 0], [1, 0],   # bottom, top (x direction)
+            [-1, -1], [-1, 1], # corners
+            [1, -1], [1, 1]
+        ]
+    else:
+        # For 3D: work with z, y, x coordinates (indices 0, 1, 2)
+        active_dims = [0, 1, 2]  # z, y, x
+        # 26 potential mirror regions for 3D (3^3 - 1, excluding center)
+        mirror_combinations = []
+        for dz in [-1, 0, 1]:
+            for dy in [-1, 0, 1]:
+                for dx in [-1, 0, 1]:
+                    if not (dz == 0 and dy == 0 and dx == 0):  # exclude center
+                        mirror_combinations.append([dz, dy, dx])
+    
+    # Process each potential mirror region
+    for mirror_dir in mirror_combinations:
+        # Find points that need this specific mirroring
+        needs_mirror = np.ones(len(points_array), dtype=bool)
+        
+        # Check each active dimension
+        for i, dim_idx in enumerate(active_dims):
+            direction = mirror_dir[i] if dim == 3 else mirror_dir[i]
+            
+            # Safety check: make sure we have bounds for this dimension
+            if dim_idx >= len(min_coords) or dim_idx >= len(max_coords):
+                needs_mirror = np.zeros(len(points_array), dtype=bool)  # Skip this mirror if bounds insufficient
+                break
+            
+            if direction == -1:  # Points near minimum boundary
+                # Distance from point to min boundary < max_r
+                needs_mirror &= (points_array[:, dim_idx] - min_coords[dim_idx]) < max_r
+            elif direction == 1:  # Points near maximum boundary  
+                # Distance from point to max boundary < max_r
+                needs_mirror &= (max_coords[dim_idx] - points_array[:, dim_idx]) < max_r
+            # direction == 0 means no constraint for this dimension
+        
+        # Create mirrored points if any qualify
+        if np.any(needs_mirror):
+            mirrored_points = points_array[needs_mirror].copy()
+            
+            # Apply mirroring transformation for each active dimension
+            for i, dim_idx in enumerate(active_dims):
+                direction = mirror_dir[i] if dim == 3 else mirror_dir[i]
+                
+                # Safety check again
+                if dim_idx >= len(min_coords) or dim_idx >= len(max_coords):
+                    continue
+                
+                if direction == -1:  # Mirror across minimum boundary
+                    # Reflection formula: new_coord = 2 * boundary - old_coord
+                    mirrored_points[:, dim_idx] = 2 * min_coords[dim_idx] - mirrored_points[:, dim_idx]
+                elif direction == 1:  # Mirror across maximum boundary
+                    # Reflection formula: new_coord = 2 * boundary - old_coord
+                    mirrored_points[:, dim_idx] = 2 * max_coords[dim_idx] - mirrored_points[:, dim_idx]
+            
+            # Add mirrored points to collection
+            all_points = np.vstack([all_points, mirrored_points])
+    
+    return all_points
+def get_max_r_from_proportion(bounds, proportion):
+    """
+    Calculate max_r based on bounds and proportion, matching your generate_r_values logic.
+    
+    Parameters:
+    bounds: tuple of (min_coords, max_coords)
+    proportion: maximum proportion of study area extent
+    
+    Returns:
+    max_r value
+    """
+    min_coords, max_coords = bounds
+    min_coords = np.array(min_coords)
+    max_coords = np.array(max_coords)
+    
+    # Calculate dimensions
+    dimensions = max_coords - min_coords
+    
+    # Remove placeholder dimensions (where dimension = 1, typically for 2D z-dimension)
+    # But ensure we don't end up with an empty array
+    filtered_dimensions = dimensions[dimensions != 1]
+    if len(filtered_dimensions) == 0:
+        # If all dimensions were 1 (shouldn't happen), use original dimensions
+        filtered_dimensions = dimensions
+    
+    # Use minimum dimension for safety (matches your existing logic)
+    min_dimension = np.min(filtered_dimensions)
+    max_r = min_dimension * proportion
+    
+    return max_r
+
+def apply_edge_correction_to_ripley(roots, targs, proportion, bounds, dim, node_centroids=None):
+    """
+    Apply edge correction through mirroring to target points.
+    
+    This should be called AFTER convert_centroids_to_array but BEFORE 
+    convert_augmented_array_to_points (for 2D case).
+    
+    Parameters:
+    roots: array of root points (search centers) - already scaled
+    targs: array of target points (points being searched for) - already scaled  
+    proportion: the proportion parameter from your workflow
+    bounds: boundary tuple (min_coords, max_coords) or None
+    dim: dimension (2 or 3)
+    node_centroids: dict of node centroids (needed if bounds is None)
+    
+    Returns:
+    tuple: (roots, mirrored_targs) where mirrored_targs includes edge corrections
+    """
+    # Handle bounds calculation if not provided (matching your existing logic)
+    if bounds is None:
+        if node_centroids is None:
+            # Fallback: calculate from the points we have
+            all_points = np.vstack([roots, targs])
+        else:
+            # Use your existing method
+            import proximity  # Assuming this is available
+            big_array = proximity.convert_centroids_to_array(list(node_centroids.values()))
+            all_points = big_array
+        
+        min_coords = np.array([0, 0, 0])
+        max_coords = [np.max(all_points[:, 0]), np.max(all_points[:, 1]), np.max(all_points[:, 2])]
+        max_coords = np.flip(max_coords)  # Convert [x,y,z] to [z,y,x] format
+        bounds = (min_coords, max_coords)
+        
+        if 'big_array' in locals():
+            del big_array
+    
+    # Calculate max_r using your existing logic
+    max_r = get_max_r_from_proportion(bounds, proportion)
+    
+    # Mirror target points for edge correction
+    mirrored_targs = mirror_points_for_edge_correction(targs, bounds, max_r, dim)
+    
+    print(f"Original target points: {len(targs)}, After mirroring: {len(mirrored_targs)}")
+    print(f"Added {len(mirrored_targs) - len(targs)} mirrored points for edge correction")
+    print(f"Using max_r = {max_r} for mirroring threshold")
+    print(f"Bounds used: min={bounds[0]}, max={bounds[1]}")
+    
+    return roots, mirrored_targs
+
 
 #CLASSLESS FUNCTIONS THAT MAY BE USEFUL TO USERS TO RUN DIRECTLY THAT SUPPORT ANALYSIS IN SOME WAY. NOTE THESE METHODS SOMETIMES ARE USED INTERNALLY AS WELL:
 
@@ -1803,13 +2009,21 @@ def label_branches(array, peaks = 0, branch_removal = 0, comp_dil = 0, max_vol =
     if nodes is None:
 
         array = smart_dilate.smart_label(array, other_array, GPU = GPU, remove_template = True)
+        #distance = smart_dilate.compute_distance_transform_distance(array)
+        print("Watershedding result...")
+        #array = water(-distance, other_array, mask=array) #Tried out skimage watershed as shown and found it did not label branches as well as smart_label (esp combined combined with post-processing label splitting if needed)
 
     else:
         if down_factor is not None:
             array = smart_dilate.smart_label(bonus_array, array, GPU = GPU, predownsample = down_factor, remove_template = True)
+            #distance = smart_dilate.compute_distance_transform_distance(bonus_array)
+            #array = water(-distance, array, mask=bonus_array)
         else:
 
             array = smart_dilate.smart_label(bonus_array, array, GPU = GPU, remove_template = True)
+            #distance = smart_dilate.compute_distance_transform_distance(bonus_array)
+            #array = water(-distance, array, mask=bonus_array)
+
 
     if down_factor is not None and nodes is None:
         array = upsample_with_padding(array, down_factor, arrayshape)
@@ -2089,6 +2303,56 @@ def filter_size_by_vol(binary_array, volume_threshold):
     
     return result
 
+def gray_watershed(image, min_distance = 1, threshold_abs = None):
+
+
+    from skimage.feature import peak_local_max
+
+    if len(np.unique(image)) == 2:
+        image = smart_dilate.compute_distance_transform_distance(image)
+
+
+    is_pseudo_3d = image.shape[0] == 1
+    if is_pseudo_3d:
+        image = np.squeeze(image)  # Convert to 2D for processing
+
+    #smoothed = ndimage.gaussian_filter(image.astype(float), sigma=2)
+
+    peaks = peak_local_max(image, min_distance = min_distance, threshold_abs = threshold_abs)
+    if len(peaks) < 256:
+        dtype = np.uint8
+    elif len(peaks) < 65535:
+        dtype = np.uint16
+    else:
+        dytpe = np.uint32
+
+    clone = np.zeros_like(image).astype(dtype)
+
+    if not is_pseudo_3d:
+        for i, peak in enumerate(peaks):
+            z, y, x = peak
+            clone[z,y,x] = i + 1
+    else:
+        for i, peak in enumerate(peaks):
+            y, x = peak
+            clone[y,x] = i + 1
+
+
+    if is_pseudo_3d:
+        image = np.expand_dims(image, axis = 0)
+        clone = np.expand_dims(clone, axis = 0)
+
+
+    binary_image = binarize(image)
+    #image = smart_dilate.smart_label(image, clone, GPU = False)
+
+    image = water(-image, clone, mask=binary_image)
+
+
+
+    return image
+
+
 def watershed(image, directory = None, proportion = 0.1, GPU = True, smallest_rad = None, predownsample = None, predownsample2 = None):
     """
     Can be used to 3D watershed a binary image. Watershedding attempts to use an algorithm to split touching objects into seperate labelled components. Labelled output will be saved to the active directory if none is specified.
@@ -2162,15 +2426,17 @@ def watershed(image, directory = None, proportion = 0.1, GPU = True, smallest_ra
     if len(labels.shape) ==2:
         labels = np.expand_dims(labels, axis = 0)
 
-    del distance
+    #del distance
 
 
     if labels.shape[1] < original_shape[1]: #If downsample was used, upsample output
         labels = upsample_with_padding(labels, downsample_needed, original_shape)
         labels = labels * old_mask
-        labels = smart_dilate.smart_label(old_mask, labels, GPU = GPU, predownsample = predownsample2)
+        labels = water(-distance, labels, mask=old_mask) # Here i like skimage watershed over smart_label, mainly because skimage just kicks out too-small nodes from the image, while smart label just labels them sort of wrongly.
+        #labels = smart_dilate.smart_label(old_mask, labels, GPU = GPU, predownsample = predownsample2)
     else:
-        labels = smart_dilate.smart_label(image, labels, GPU = GPU, predownsample = predownsample2)
+        labels = water(-distance, labels, mask=image)
+        #labels = smart_dilate.smart_label(image, labels, GPU = GPU, predownsample = predownsample2)
 
     if directory is None:
         pass
@@ -3389,8 +3655,6 @@ class Network_3D:
         if skeletonized:
             binary_edges = skeletonize(binary_edges)
 
-
-
         if search is not None and hasattr(self, '_nodes') and self._nodes is not None and self._search_region is None:
             search_region = binarize(self._nodes)
             dilate_xy, dilate_z = dilation_length_to_pixels(self._xy_scale, self._z_scale, search, search)
@@ -3550,11 +3814,12 @@ class Network_3D:
             self._network_lists = network_analysis.read_excel_to_lists(df)
             self._network, net_weights = network_analysis.weighted_network(df)
 
-        if ignore_search_region and hasattr(self, '_edges') and self._edges is not None and hasattr(self, '_nodes') and self._nodes is not None and search is not None:
-            dilate_xy, dilate_z = dilation_length_to_pixels(self._xy_scale, self._z_scale, search, search)
-            print(f"{dilate_xy}, {dilate_z}")
+        if ignore_search_region and hasattr(self, '_edges') and self._edges is not None and hasattr(self, '_nodes') and self._nodes is not None:
+            #dilate_xy, dilate_z = dilation_length_to_pixels(self._xy_scale, self._z_scale, search, search)
+            #print(f"{dilate_xy}, {dilate_z}")
             num_nodes = np.max(self._nodes)
-            connections_parallel = create_node_dictionary(self._nodes, self._edges, num_nodes, dilate_xy, dilate_z) #Find which edges connect which nodes and put them in a dictionary.
+            #connections_parallel = create_node_dictionary(self._nodes, self._edges, num_nodes, dilate_xy, dilate_z) #Find which edges connect which nodes and put them in a dictionary.
+            connections_parallel = create_node_dictionary(self._nodes, self._edges, num_nodes, 3, 3) #For now I only implement this for immediate neighbor search so we'll just use 3 and 3 here.
             connections_parallel = find_shared_value_pairs(connections_parallel) #Sort through the dictionary to find connected node pairs.
             df = create_and_save_dataframe(connections_parallel)
             self._network_lists = network_analysis.read_excel_to_lists(df)
@@ -3635,7 +3900,10 @@ class Network_3D:
                 except:
                     pass
 
-        self.calculate_edges(edges, diledge = diledge, inners = inners, hash_inner_edges = hash_inners, search = search, remove_edgetrunk = remove_trunk, GPU = GPU, fast_dil = fast_dil, skeletonized = skeletonize)
+            self.calculate_edges(edges, diledge = diledge, inners = inners, hash_inner_edges = hash_inners, search = search, remove_edgetrunk = remove_trunk, GPU = GPU, fast_dil = fast_dil, skeletonized = skeletonize) #Will have to be moved out if the second method becomes more directly implemented
+        else:
+            self._edges, _ = label_objects(edges)
+
         del edges
         if directory is not None:
             try:
@@ -4650,27 +4918,50 @@ class Network_3D:
         return neighborhood_dict, proportion_dict, title1, title2, densities
 
 
-    def get_ripley(self, root = None, targ = None, distance = 1, edgecorrect = True, bounds = None, ignore_dims = False, proportion = 0.5):
+
+    def get_ripley(self, root = None, targ = None, distance = 1, edgecorrect = True, bounds = None, ignore_dims = False, proportion = 0.5, mode = 0, safe = False, factor = 0.25):
+
+        is_subset = False
+
+        if bounds is None:
+            big_array = proximity.convert_centroids_to_array(list(self.node_centroids.values()))
+            min_coords = np.array([0,0,0])
+            max_coords = [np.max(big_array[:, 0]), np.max(big_array[:, 1]), np.max(big_array[:, 2])]
+            del big_array
+            max_coords = np.flip(max_coords)
+            bounds = (min_coords, max_coords)
+        else:
+            min_coords, max_coords = bounds
+
+        min_bounds, max_bounds = bounds
+        sides = max_bounds - min_bounds
+        # Set max_r to None since we've handled edge effects through mirroring
 
 
         if root is None or targ is None: #Self clustering in this case
             roots = self._node_centroids.values()
+            root_ids = self.node_centroids.keys()
             targs = self._node_centroids.values()
+            is_subset = True
         else:
             roots = []
             targs = []
+            root_ids = []
 
             for node, nodeid in self.node_identities.items(): #Otherwise we need to pull out this info
                 if nodeid == root:
                     roots.append(self._node_centroids[node])
+                    root_ids.append(node)
                 if nodeid == targ:
                     targs.append(self._node_centroids[node])
 
+        if not is_subset:
+            if np.array_equal(roots, targs):
+                is_subset = True
+
         rooties = proximity.convert_centroids_to_array(roots, xy_scale = self.xy_scale, z_scale = self.z_scale)
         targs = proximity.convert_centroids_to_array(targs, xy_scale = self.xy_scale, z_scale = self.z_scale)
-        points_array = np.vstack((rooties, targs))
-        del rooties
-
+        
         try:
             if self.nodes.shape[0] == 1:
                 dim = 2
@@ -4683,66 +4974,135 @@ class Network_3D:
                     dim = 3
                     break
 
+        if dim == 2:
+            volume = sides[0] * sides[1] * self.xy_scale**2
+        else:
+            volume = np.prod(sides) * self.z_scale * self.xy_scale**2
+
+        points_array = np.vstack((rooties, targs))
+        del rooties
+        max_r = None
+        if safe:
+            proportion = factor
+            
+
         if ignore_dims:
 
-            factor = 0.25
+            new_list = []
 
-            big_array = proximity.convert_centroids_to_array(list(self.node_centroids.values()))
+            if mode == 0:
 
-            if bounds is None:
-                min_coords = np.array([0,0,0])
-                max_coords = [np.max(big_array[:, 0]), np.max(big_array[:, 1]), np.max(big_array[:, 2])]
-                del big_array
-                max_coords = np.flip(max_coords)
-                bounds = (min_coords, max_coords)
-            else:
-                min_coords, max_coords = bounds
+                try:
+                    dim_list = max_coords - min_coords
+                except:
+                    min_coords = np.array([0,0,0])
+                    bounds = (min_coords, max_coords)
+                    dim_list = max_coords - min_coords
 
-            try:
-                dim_list = max_coords - min_coords
-            except:
-                min_coords = np.array([0,0,0])
-                bounds = (min_coords, max_coords)
-                dim_list = max_coords - min_coords
 
-            new_list = []
+                if dim == 3:
 
+                    """
+                    if self.xy_scale > self.z_scale: # xy will be 'expanded' more so its components will be arbitrarily further from the border than z ones
+                        factor_xy = (self.z_scale/self.xy_scale) * factor # So 'factor' in the xy dim has to get smaller
+                        factor_z = factor
+                    elif self.z_scale > self.xy_scale: # Same idea
+                        factor_z = (self.xy_scale/self.z_scale) * factor
+                        factor_xy = factor
+                    else:
+                        factor_z = factor
+                        factor_xy = factor
+                    """
 
-            if dim == 3:
-                for centroid in roots:
+                    for centroid in roots:
 
-                    if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor) and ((centroid[0] - min_coords[2]) > dim_list[2] * factor) and ((max_coords[2] - centroid[0]) > dim_list[2] * factor):
-                        new_list.append(centroid)
+                        if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor) and ((centroid[0] - min_coords[2]) > dim_list[2] * factor) and ((max_coords[2] - centroid[0]) > dim_list[2] * factor):
+                            new_list.append(centroid)
 
 
-                    #if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor) and ((centroid[0] - min_coords[2]) > dim_list[2] * factor) and ((max_coords[2] - centroid[0]) > dim_list[2] * factor):
-                        #new_list.append(centroid)
-                        #print(f"dim_list: {dim_list}, centroid: {centroid}, min_coords: {min_coords}, max_coords: {max_coords}")
+                        #if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor) and ((centroid[0] - min_coords[2]) > dim_list[2] * factor) and ((max_coords[2] - centroid[0]) > dim_list[2] * factor):
+                            #new_list.append(centroid)
+                            #print(f"dim_list: {dim_list}, centroid: {centroid}, min_coords: {min_coords}, max_coords: {max_coords}")
+                else:
+                    for centroid in roots:
+
+                        if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor):
+                            new_list.append(centroid)
+
             else:
-                for centroid in roots:
 
-                    if ((centroid[2] - min_coords[0]) > dim_list[0] * factor) and ((max_coords[0] - centroid[2]) > dim_list[0]  * factor) and ((centroid[1] - min_coords[1]) > dim_list[1] * factor) and ((max_coords[1] - centroid[1]) > dim_list[1] * factor):
-                        new_list.append(centroid)
+                if mode == 1:
+
+                    legal = self.edges != 0
+
+                elif mode == 2:
+
+                    legal = self.network_overlay != 0
+
+                elif mode == 3:
+
+                    legal = self.id_overlay != 0
+
+                if self.nodes is None:
+
+                    temp_array = proximity.populate_array(self.node_centroids, shape = legal.shape)
+                else:
+                    temp_array = self.nodes
+
+                if dim == 2:
+                    volume = np.count_nonzero(legal) * self.xy_scale**2
+                else:
+                    volume = np.count_nonzero(legal) * self.z_scale * self.xy_scale**2
+                print(f"Using {volume} for the volume measurement (Volume of provided mask as scaled by xy and z scaling)")
+
+                legal = smart_dilate.compute_distance_transform_distance(legal, sampling = [self.z_scale, self.xy_scale, self.xy_scale]) # Get true distances
+                
+                max_avail = np.max(legal) # Most internal point
+                min_legal = factor * max_avail # Values of stuff 25% within the tissue
+
+                legal = legal > min_legal
+
+                if safe:
+                    max_r = min_legal
+
+
+                legal = temp_array * legal
+
+                legal = np.unique(legal)
+                if 0 in legal:
+                    legal = np.delete(legal, 0)
+                for node in legal:
+                    if node in root_ids:
+                        new_list.append(self.node_centroids[node])
 
             roots = new_list
             print(f"Utilizing {len(roots)} root points. Note that low n values are unstable.")
             is_subset = True
-        else:
-            is_subset = False
-
 
 
 
         roots = proximity.convert_centroids_to_array(roots, xy_scale = self.xy_scale, z_scale = self.z_scale)
 
+        n_subset = len(targs)
+
+        # Apply edge correction through mirroring
+        if edgecorrect:
+
+
+            roots, targs = apply_edge_correction_to_ripley(
+                roots, targs, proportion, bounds, dim, 
+                node_centroids=self.node_centroids  # Pass this for bounds calculation if needed
+            )
+
 
         if dim == 2:
             roots = proximity.convert_augmented_array_to_points(roots)
             targs = proximity.convert_augmented_array_to_points(targs)
 
-        r_vals = proximity.generate_r_values(points_array, distance, bounds = bounds, dim = dim, max_proportion=proportion)
+        print(f"Using {len(roots)} root points")
+        r_vals = proximity.generate_r_values(points_array, distance, bounds = bounds, dim = dim, max_proportion=proportion, max_r = max_r)
 
-        k_vals =  proximity.optimized_ripleys_k(roots, targs, r_vals, bounds=bounds, edge_correction=edgecorrect, dim = dim, is_subset = is_subset)
+        k_vals =  proximity.optimized_ripleys_k(roots, targs, r_vals, bounds=bounds, dim = dim, is_subset = is_subset, volume = volume, n_subset = n_subset)
 
         h_vals = proximity.compute_ripleys_h(k_vals, r_vals, dim)
 
@@ -4792,7 +5152,7 @@ class Network_3D:
             search_x, search_z = dilation_length_to_pixels(self._xy_scale, self._z_scale, search, search)
 
 
-        num_nodes = np.max(self._nodes)
+        num_nodes = int(np.max(self._nodes))
 
         my_dict = proximity.create_node_dictionary(self._nodes, num_nodes, search_x, search_z, targets = targets, fastdil = fastdil, xy_scale = self._xy_scale, z_scale = self._z_scale, search = search)
 
@@ -4804,7 +5164,7 @@ class Network_3D:
 
         self.remove_edge_weights()
 
-    def centroid_array(self, clip = False):
+    def centroid_array(self, clip = False, shape = None):
         """Use the centroids to populate a node array"""
 
         if clip:
@@ -4814,12 +5174,13 @@ class Network_3D:
 
         else:
 
-            array = proximity.populate_array(self.node_centroids)
+            array = proximity.populate_array(self.node_centroids, shape = shape)
 
             return array
 
 
 
+
     def random_nodes(self, bounds = None, mask = None):
 
         if self.nodes is not None:
@@ -4938,6 +5299,13 @@ class Network_3D:
 
         return output
 
+    def centroid_umap(self):
+
+        from . import neighborhoods
+
+        neighborhoods.visualize_cluster_composition_umap(self.node_centroids, None, id_dictionary = self.node_identities, graph_label = "Node ID", title = 'UMAP Visualization of Node Centroids') 
+
+
     def community_id_info_per_com(self, umap = False, label = False, limit = 0, proportional = False):
 
         community_dict = invert_dict(self.communities)
@@ -5043,21 +5411,23 @@ class Network_3D:
 
         zero_group = {}
 
+        comus = invert_dict(self.communities)
 
-        if limit is not None:
-
-            coms = invert_dict(self.communities)
 
+        if limit is not None:
 
-            for com, nodes in coms.items():
+            for com, nodes in comus.items():
 
                 if len(nodes) < limit:
 
                     del identities[com]
 
-        if count > len(identities):
-            print(f"Requested neighborhoods too large for available communities. Using {len(identities)} neighborhoods (max for these coms)")
-            count = len(identities)
+        try:
+            if count > len(identities):
+                print(f"Requested neighborhoods too large for available communities. Using {len(identities)} neighborhoods (max for these coms)")
+                count = len(identities)
+        except:
+            pass
 
 
         if mode == 0:
@@ -5068,13 +5438,21 @@ class Network_3D:
         coms = {}
 
         neighbors = {}
+        len_dict = {}
+        inc_count = 0
 
         for i, cluster in enumerate(clusters):
 
+            size = len(cluster)
+            inc_count += size
+
+            len_dict[i + 1] = [size]
+
             for com in cluster: # For community ID per list
 
                 coms[com] = i + 1
 
+
         copy_dict = copy.deepcopy(self.communities)
 
         for node, com in copy_dict.items():
@@ -5092,18 +5470,17 @@ class Network_3D:
 
         if len(zero_group) > 0:
             self.communities.update(zero_group)
+            len_dict[0] = [len(comus) - inc_count]
 
 
         identities, id_set = self.community_id_info_per_com()
 
-        len_dict = {}
-
         coms = invert_dict(self.communities)
         node_count = len(list(self.communities.keys()))
 
         for com, nodes in coms.items():
 
-            len_dict[com] = len(nodes)/node_count
+            len_dict[com].append(len(nodes)/node_count)
 
         matrixes = []
 
@@ -5119,7 +5496,7 @@ class Network_3D:
 
             identities3 = {}
             for iden in identities2:
-                identities3[iden] = identities2[iden]/len_dict[iden]
+                identities3[iden] = identities2[iden]/len_dict[iden][1]
 
             output = neighborhoods.plot_dict_heatmap(identities3, id_set2, title = "Neighborhood Heatmap by Proportional Composition of Nodes in Neighborhood vs All Nodes Divided by Neighborhood Total Proportion of All Nodes (val < 1 = underrepresented, val > 1 = overrepresented)", center_at_one = True)
             matrixes.append(output)
@@ -5291,7 +5668,33 @@ class Network_3D:
                     pass
 
 
-    def nearest_neighbors_avg(self, root, targ, xy_scale = 1, z_scale = 1, num = 1, heatmap = False, threed = True, numpy = False):
+    def nearest_neighbors_avg(self, root, targ, xy_scale = 1, z_scale = 1, num = 1, heatmap = False, threed = True, numpy = False, quant = False):
+
+        def get_theoretical_nearest_neighbor_distance(compare_set, num_neighbors, volume, is_2d=False):
+            """
+            Calculate theoretical expected distance to k-th nearest neighbor 
+            assuming random uniform distribution in 2D or 3D space.
+            """
+            import math
+            
+            if len(compare_set) == 0 or volume <= 0:
+                raise ValueError("Invalid input: empty set or non-positive volume")
+            
+            density = len(compare_set) / volume
+            k = num_neighbors
+            
+            if is_2d:
+                # Expected distance to k-th nearest neighbor in 2D
+                # μ1' = Γ(k + 1/2) / (Γ(k) × √(m × π))
+                expected_distance = math.gamma(k + 0.5) / (math.gamma(k) * math.sqrt(density * math.pi))
+            else:
+                # Expected distance to k-th nearest neighbor in 3D
+                # μ1' = Γ(k + 1/3) / (Γ(k) × (m × Φ)^(1/3))
+                # where Φ = π^(3/2) / Γ(3/2 + 1) = π^(3/2) / Γ(5/2) = 4π/3
+                phi_3d = 4 * math.pi / 3  # Volume of unit sphere in 3D
+                expected_distance = math.gamma(k + 1/3) / (math.gamma(k) * (density * phi_3d)**(1/3))
+            
+            return expected_distance
 
         root_set = []
 
@@ -5335,6 +5738,14 @@ class Network_3D:
             
         avg, output = proximity.average_nearest_neighbor_distances(self.node_centroids, root_set, compare_set, xy_scale=self.xy_scale, z_scale=self.z_scale, num = num)
 
+        if quant:
+            try:
+                quant_overlay = node_draw.degree_infect(output, self._nodes, make_floats = True)
+            except:
+                quant_overlay = None
+        else:
+            quant_overlay = None
+
         if heatmap:
 
 
@@ -5345,7 +5756,30 @@ class Network_3D:
                 big_array = proximity.convert_centroids_to_array(list(self.node_centroids.values()))
                 shape = [np.max(big_array[0, :]) + 1, np.max(big_array[1, :]) + 1, np.max(big_array[2, :]) + 1]
 
-            pred = avg
+
+            try:
+                bounds = self.nodes.shape
+            except:
+                try:
+                    bounds = self.edges.shape
+                except:
+                    try:
+                        bounds = self.network_overlay.shape
+                    except:
+                        try:
+                            bounds = self.id_overlay.shape
+                        except:
+                            big_array = proximity.convert_centroids_to_array(list(self.node_centroids.values()))
+                            max_coords = [np.max(big_array[:, 0]), np.max(big_array[:, 1]), np.max(big_array[:, 2])]
+                            del big_array
+            volume = bounds[0] * bounds[1] * bounds[2] * self.z_scale * self.xy_scale**2
+            if 1 in bounds or 0 in bounds:
+                is_2d = True
+            else:
+                is_2d = False
+
+            pred = get_theoretical_nearest_neighbor_distance(compare_set, num, volume, is_2d = is_2d)
+            #pred = avg
 
             node_intensity = {}
             import math
@@ -5359,12 +5793,12 @@ class Network_3D:
 
                 overlay = neighborhoods.create_node_heatmap(node_intensity, node_centroids, shape = shape, is_3d=threed, labeled_array = self.nodes, colorbar_label="Clustering Intensity", title = title)
 
-                return avg, output, overlay
+                return avg, output, overlay, quant_overlay
 
             else:
                 neighborhoods.create_node_heatmap(node_intensity, node_centroids, shape = shape, is_3d=threed, labeled_array = None, colorbar_label="Clustering Intensity", title = title)
 
-        return avg, output
+        return avg, output, quant_overlay