Merge pull request #350 from w-k-jones/PBC_feature_location_fix

PBC feature location fix

.readthedocs.yml

@@ -1,6 +1,9 @@
 version: 2
 formats: all
+build:
+  os: ubuntu-22.04
+  tools:
+    python: "3.9"
 python:
-  version: 3
   install:
-    - requirements: doc/requirements.txt
+    - requirements: doc/requirements.txt

doc/requirements.txt

@@ -1,3 +1,4 @@
 ipykernel
 nbsphinx
 numpy
+sphinx_rtd_theme

tobac/segmentation.py

@@ -104,8 +104,13 @@ def add_markers(
         marker_arr = marker_arr[np.newaxis, :, :]
 
     if seed_3D_flag == "column":
-        for index, row in features.iterrows():
-            marker_arr[level, int(row["hdim_1"]), int(row["hdim_2"])] = row["feature"]
+        for _, row in features.iterrows():
+            # Offset marker locations by 0.5 to find nearest pixel
+            marker_arr[
+                level,
+                int(row["hdim_1"] + 0.5) % h1_len,
+                int(row["hdim_2"] + 0.5) % h2_len,
+            ] = row["feature"]
 
     elif seed_3D_flag == "box":
         # Get the size of the seed box from the input parameter
@@ -123,7 +128,7 @@ def add_markers(
             seed_h1 = seed_3D_size
             seed_h2 = seed_3D_size
 
-        for index, row in features.iterrows():
+        for _, row in features.iterrows():
             if is_3D:
                 # If we have a 3D input and we need to do box seeding
                 # we need to have 3D features.
@@ -781,7 +786,7 @@ def segmentation_timestep(
                 if buddy == 0:
                     continue
                 # isolate feature from set of buddies
-                buddy_feat = features_in[features_in["feature"] == buddy]
+                buddy_feat = features_in[features_in["feature"] == buddy].iloc[0]
 
                 # transform buddy feature position if needed for positioning in z2/y2/x2 space
                 # MAY be redundant with what is done just below here

tobac/utils/periodic_boundaries.py

@@ -297,7 +297,10 @@ def weighted_circmean(
         angle_average = np.arctan2(sin_average, cos_average) % (2 * np.pi)
     rescaled_average = (angle_average * scaling_factor) + low
     # Round return value to try and supress rounding errors
-    return np.round(rescaled_average, 12)
+    rescaled_average = np.round(rescaled_average, 12)
+    if rescaled_average == high:
+        rescaled_average = low
+    return rescaled_average
 
 
 def transfm_pbc_point(in_dim, dim_min, dim_max):