add df

src/sparcscore/pipeline/workflows.py

@@ -19,6 +19,7 @@
 import os
 import sys
 import numpy as np
+import pandas as pd
 import torch
 import gc
 import matplotlib.pyplot as plt
@@ -811,16 +812,21 @@ def cellpose_segmentation(self, input_image):
             masks_cytosol_unfiltered = masks_cytosol.copy()
 
         #log start time of cell filtering to track
+        timing_info = []
+
         start = time.time()
 
+        timing_info.append(("1. Time when started the whole approach", start))
+
         all_nucleus_ids = np.unique(masks_nucleus)[1:]
         nucleus_cytosol_pairs = {}
 
-        self.log(f"1. Number of nuclei to filter: {len(all_nucleus_ids)}")
+        self.log(f"Number of nuclei to filter: {len(all_nucleus_ids)}")
 
         current_time = time.time()
 
-        self.log("2. Started filtering cells (for nucleus_id in all_nucleus_ids).")
+        timing_info.append(("2. Time when started filtering cells (for nucleus_id in all_nucleus_ids)", current_time))
+
         for nucleus_id in all_nucleus_ids:
             time_in_the_loop = time.time()
 
@@ -829,14 +835,14 @@ def cellpose_segmentation(self, input_image):
             # now get the coordinates of the nucleus
             nucleus_pixels = np.nonzero(nucleus)
 
-            self.log(f"3. Time required for getting nucleus pixels in seconds: {time.time() - time_in_the_loop}")
+            timing_info.append((f"3. Time required for getting nucleus pixels in seconds for nucleus {nucleus_id}", time.time() - time_in_the_loop))
 
             time_in_the_loop = time.time()
 
             # check if those indices are not background in the cytosol mask
             potential_cytosol = masks_cytosol[nucleus_pixels]
 
-            self.log(f"4. Time required for getting potential cytosol pixels in seconds: {time.time() - time_in_the_loop}")
+            timing_info.append((f"4. Time required for getting potential cytosol pixels in seconds for nucleus {nucleus_id}", time.time() - time_in_the_loop))
 
             if np.all(potential_cytosol != 0):
                 time_in_the_loop = time.time()
@@ -847,7 +853,7 @@ def cellpose_segmentation(self, input_image):
                 all_counts = np.sum(counts)
                 cytosol_proportions = counts / all_counts
 
-                self.log(f"5. Time required for getting unique cytosol pixels and calculating their proportions in seconds: {time.time() - time_in_the_loop}")
+                timing_info.append((f"5. Time required for getting unique cytosol pixels and calculating their proportions in seconds for nucleus {nucleus_id}", time.time() - time_in_the_loop))
 
                 if np.any(cytosol_proportions >= self.config["filtering_threshold"]):
                     time_in_the_loop = time.time()
@@ -860,31 +866,31 @@ def cellpose_segmentation(self, input_image):
                 else:
                     nucleus_cytosol_pairs[nucleus_id] = 0
 
-                self.log(f"6. Time required for getting cytosol_id with max proportion in seconds: {time.time() - time_in_the_loop}")
+                timing_info.append((f"6. Time required for getting cytosol_id with max proportion in seconds for nucleus {nucleus_id}", time.time() - time_in_the_loop))
 
-        self.log("7. Time required for filtering cells (for nucleus_id in all_nucleus_ids) in seconds: {}".format(time.time() - current_time))
+        timing_info.append(("7. Time required for filtering cells (for nucleus_id in all_nucleus_ids) in seconds", time.time() - current_time))
 
         # check if there are any cytosol masks that are assigned to multiple nuclei
         cytosol_count = defaultdict(int)
 
-        self.log("8. Started filtering cytosol masks assigned to multiple nuclei.")
-
         new_time = time.time()
 
+        timing_info.append(("8. Time when started counting the occurences of each cytosol id", new_time))
+
         # Count the occurrences of each cytosol value
         for cytosol in nucleus_cytosol_pairs.values():
             cytosol_count[cytosol] += 1
 
-        self.log("9. Time required for counting the occurences of each cytosol id in seconds: {}".format(time.time() - new_time))
+        timing_info.append(("9. Time required for counting the occurences of each cytosol id in seconds", time.time() - new_time))
 
         new_time = time.time()
 
         # Find cytosol values assigned to more than one nucleus
         for nucleus, cytosol in nucleus_cytosol_pairs.items():
             if cytosol_count[cytosol] > 1:
                 nucleus_cytosol_pairs[nucleus] = 0
-            
-        self.log("10. Time required for counting cytosol masks in seconds: {}".format(time.time() - new_time))
+
+        timing_info.append(("10. Time required for filtering cytosol ids that are assigned to more than one nucleus in seconds", time.time() - new_time))
 
         new_time = time.time()
 
@@ -898,7 +904,7 @@ def cellpose_segmentation(self, input_image):
         for cytosol_id in not_used_cytosol_ids:
             masks_cytosol[masks_cytosol == cytosol_id] = 0
 
-        self.log("11. Time required for filtering cytosol masks that are not in the lookup table in seconds: {}".format(time.time() - new_time))
+        timing_info.append(("11. Time required for filtering cytosol masks that are not in the lookup table in seconds", time.time() - new_time))
 
         new_time = time.time()
 
@@ -911,8 +917,8 @@ def cellpose_segmentation(self, input_image):
         # set all nucleus ids that are not present in lookup table to 0 in the nucleus mask
         for nucleus_id in not_used_nucleus_ids:
             masks_nucleus[masks_nucleus == nucleus_id] = 0
-
-        self.log("12. Time required for filtering nucleus masks that are not in the lookup table in seconds: {}".format(time.time() - new_time))
+        
+        timing_info.append(("12. Time required for filtering nucleus masks that are not in the lookup table in seconds", time.time() - new_time))
 
         new_time = time.time()
 
@@ -929,11 +935,15 @@ def cellpose_segmentation(self, input_image):
                 masks_cytosol[condition] = nucleus_id
                 updated_cytosol_mask = np.logical_or(updated_cytosol_mask, condition)
 
-        self.log("13. Time required for updating nucleus-cytosol pairs in seconds: {}".format(time.time() - new_time))
+        timing_info.append(("13. Time required for filtering cytosol masks that are not in the lookup table in seconds", time.time() - new_time))
 
         end = time.time()
+
+        timing_info.append(("14. Time required for filtering generated masks in seconds", end - start))
 
         self.log(f"Time required for filtering generated masks in seconds: {end - start}")
+
+        df_timing = pd.DataFrame(timing_info, columns=["Step", "Time (s)"])
 
         if self.debug:
             # plot nucleus and cytosol masks before and after filtering