bugs correction

brick/component/extension.py

@@ -192,6 +192,7 @@ class extension_t(glial_cmp_t):
         result = image.copy()
         if (low is not None) and (high is not None):
             result = __HysterisisImage__(result, low, high)
+            # np_.save("D:\\MorganeNadal\\img512enhanced_norm_hyst_2.npy", result)
         if selem is not None:
             result = __MorphologicalCleaning__(result, selem)
 

brick/processing/graph_feat_extraction.py

@@ -140,7 +140,7 @@ def GetNodesCoordinates(node_coord: Tuple[str, ...]) -> list:
 
     for node in range(len(node_coord)):
         coord_node = coord[node]
-        pattern = '\d+'
+        pattern = r"\d+"
         coord_node = re_.findall(pattern, coord_node)
         coor = []
         for i in range(3):
@@ -151,7 +151,7 @@ def GetNodesCoordinates(node_coord: Tuple[str, ...]) -> list:
     return coord
 
 
-def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int, max_range: float, hist_min_length: float, scale_map: array_t, decimals: int = 4):
+def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, bins_length:array_t, bins_curvature:array_t, scale_map: array_t, decimals: int = 4):
     """
     Extract the features from somas and graphs.
     Returns a pandas dataframe.
@@ -200,6 +200,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
         "median_curvature",
         "std_curvature",
         "entropy_curvature",
+        "hist_curvature",
         "min_torsion",
         "max_torsion",
         "mean_torsion",
@@ -233,6 +234,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
         "median_curvature_P",
         "std_curvature_P",
         "entropy_curvature_P",
+        "hist_curvature_P",
         "min_torsion_P",
         "max_torsion_P",
         "mean_torsion_P",
@@ -266,6 +268,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
         "median_curvature_S",
         "std_curvature_S",
         "entropy_curvature_S",
+        "hist_curvature_S",
         "min_torsion_S",
         "max_torsion_S",
         "mean_torsion_S",
@@ -320,7 +323,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             total_ext_length = in_.ToMicron(soma.skl_graph.length, size_voxel_in_micron, decimals=decimals)
             #
             # Lengths histogram
-            hist_lengths = np_.histogram(ext_lengths, bins=number_of_bins, range=(hist_min_length, max_range))[0]
+            hist_lengths = np_.histogram(ext_lengths, bins_length)[0]
             #
             # min, mean, median, max and standard deviation of the ALL extensions
             min_length = in_.ToMicron(soma.skl_graph.min_length, size_voxel_in_micron, decimals=decimals)
@@ -360,13 +363,14 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature = np_.median(curvatures)
             std_curvature = np_.std(curvatures)
             entropy_curvature = st_.entropy(curvatures)
+            hist_curvature = np_.histogram(curvatures, bins_curvature)[0]
             #
             min_torsion = min(torsions)
             max_torsion = max(torsions)
             mean_torsion = np_.mean(torsions)
             median_torsion = np_.median(torsions)
             std_torsion = np_.std(torsions)
-            if any(torsions > 0):
+            if any(torsions) > 0:
                 entropy_torsion = st_.entropy(torsions)
             else:
                 entropy_torsion = 0
@@ -386,7 +390,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             total_ext_length_P = sum(ext_lengths_P)
             #
             # Lengths histogram
-            hist_lengths_P = np_.histogram(ext_lengths_P, bins=number_of_bins, range=(hist_min_length, max_range))[0]
+            hist_lengths_P = np_.histogram(ext_lengths_P, bins_length)[0]
             #
             # min, mean, median, max and standard deviation of the PRIMARY extensions
             min_length_P = min(ext_lengths_P)
@@ -424,13 +428,14 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature_P = np_.median(curvatures_P)
             std_curvature_P = np_.std(curvatures_P)
             entropy_curvature_P = st_.entropy(curvatures_P)
+            hist_curvature_P = np_.histogram(curvatures_P, bins_curvature)[0]
             #
             min_torsion_P = min(torsions_P)
             max_torsion_P = max(torsions_P)
             mean_torsion_P = np_.mean(torsions_P)
             median_torsion_P = np_.median(torsions_P)
             std_torsion_P = np_.std(torsions_P)
-            if any(torsions_P > 0):
+            if any(torsions_P) > 0:
                 entropy_torsion_P = st_.entropy(torsions_P)
             else:
                 entropy_torsion_P = 0
@@ -458,7 +463,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
                 total_ext_length_S = sum(ext_lengths_S)
                 #
                 # Lengths histogram
-                hist_lengths_S = np_.histogram(ext_lengths_S, bins=number_of_bins, range=(hist_min_length, max_range))[0]
+                hist_lengths_S = np_.histogram(ext_lengths_S, bins_length)[0]
                 #
                 # min, mean, median, max and standard deviation of the PRIMARY extensions
                 min_length_S = min(ext_lengths_S)
@@ -494,13 +499,14 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
                 median_curvature_S = np_.median(curvatures_S)
                 std_curvature_S = np_.std(curvatures_S)
                 entropy_curvature_S = st_.entropy(curvatures_S)
+                hist_curvature_S = np_.histogram(ext_lengths_S, bins_curvature)[0]
                 #
                 min_torsion_S = min(torsions_S)
                 max_torsion_S = max(torsions_S)
                 mean_torsion_S = np_.mean(torsions_S)
                 median_torsion_S = np_.median(torsions_S)
                 std_torsion_S = np_.std(torsions_S)
-                if any(torsions_S > 0):
+                if any(torsions_S) > 0:
                     entropy_torsion_S = st_.entropy(torsions_S)
                 else:
                     entropy_torsion_S = 0
@@ -551,6 +557,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
                 median_curvature_S = -1
                 std_curvature_S = 0
                 entropy_curvature_S = 0
+                hist_curvature_S = 0
                 #
                 min_torsion_S = -1
                 max_torsion_S = -1
@@ -592,6 +599,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature = -1
             std_curvature = 0
             entropy_curvature = 0
+            hist_curvature = 0
             min_torsion = -1
             max_torsion= -1
             mean_torsion = -1
@@ -625,6 +633,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature_P = -1
             std_curvature_P = 0
             entropy_curvature_P = 0
+            hist_curvature_P = 0
             min_torsion_P = -1
             max_torsion_P = -1
             mean_torsion_P = -1
@@ -658,6 +667,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature_S = -1
             std_curvature_S = 0
             entropy_curvature_S = 0
+            hist_curvature_S = 0
             min_torsion_S = -1
             max_torsion_S = -1
             mean_torsion_S = -1
@@ -714,6 +724,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature,
             std_curvature,
             entropy_curvature,
+            hist_curvature,
             min_torsion,
             max_torsion,
             mean_torsion,
@@ -747,6 +758,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature_P,
             std_curvature_P,
             entropy_curvature_P,
+            hist_curvature_P,
             min_torsion_P,
             max_torsion_P,
             mean_torsion_P,
@@ -780,6 +792,7 @@ def ExtractFeaturesInDF(somas, size_voxel_in_micron: list, number_of_bins: int,
             median_curvature_S,
             std_curvature_S,
             entropy_curvature_S,
+            hist_curvature_S,
             min_torsion_S,
             max_torsion_S,
             mean_torsion_S,

nutrimorph.py

@@ -96,8 +96,11 @@ frangi_c = None
 bright_on_dark = None
 method = None
 hist_min_length = None
-hist_step = None
-number_of_bins = None
+hist_step_length = None
+number_of_bins_length = None
+hist_min_curvature = None
+hist_step_curvature = None
+number_of_bins_curvature = None
 with_plot = None
 in_parallel = None
 
@@ -126,12 +129,14 @@ image = in_.ImageVerification(image, channel)
 image = image[:, 512:, 512:]  # 512 # 562 # Just for development
 img_shape = image.shape
 
+# np_.save("D:\\MorganeNadal\\img512.npy", image)
 #
 print(f"IMAGE: s.{img_shape} t.{image.dtype} m.{image.min()} M.{image.max()}")
 
 # Intensity relative normalization (between 0 and 1).
 image_for_soma = in_.IntensityNormalizedImage(image)
 image_for_ext = in_.IntensityNormalizedImage(image)
+# np_.save("D:\\MorganeNadal\\img512norm.npy", image_for_ext)
 
 print(f"NRM-IMG: t.{image_for_soma.dtype} m.{image_for_soma.min():.2f} M.{image_for_soma.max():.2f}")
 
@@ -208,10 +213,13 @@ enhanced_ext, ext_scales = extension_t.EnhancedForDetection(
     method,
     in_parallel=in_parallel)
 
+# np_.save("D:\\MorganeNadal\\img512enhanced.npy", enhanced_ext)
+
 elapsed_time = tm_.gmtime(tm_.time() - start_time)
 print(f"Elapsed Time={tm_.strftime('%Hh %Mm %Ss', elapsed_time)}\n")
 
 enhanced_ext = in_.IntensityNormalizedImage(enhanced_ext)
+# np_.save("D:\\MorganeNadal\\img512enhanced_norm.npy", enhanced_ext)
 
 # Creation of the enhanced maps
 ext_nfo["coarse_map"] = extension_t.CoarseMap(enhanced_ext, ext_low_c, ext_high_c, ext_selem_pixel_c) # seuillage
@@ -223,6 +231,8 @@ ext_nfo["lmp"], n_extensions = ms_.label(ext_nfo["map"], return_num=True)
 # ext_nfo["lmp"] = relabel_sequential(ext_lmp)[0]
 # n_extensions = ext_nfo["lmp"].max()
 
+# np_.save("D:\\MorganeNadal\\img512enhanced_norm_hyst_morpho_2.npy", ext_nfo["lmp"])
+
 extensions = tuple(
     extension_t().FromMap(ext_nfo["lmp"], ext_scales, uid)
     for uid in range(1, n_extensions + 1))
@@ -362,14 +372,16 @@ print(f"\nElapsed Time={tm_.strftime('%Hh %Mm %Ss', elapsed_time)}")
 if with_plot:
     pl_.show()
 
+# np_.save("D:\\MorganeNadal\\img512final_2.npy", som_nfo['soma_w_ext_lmp'])
+
 po_.MaximumIntensityProjectionZ(som_nfo['soma_w_ext_lmp'])
-# po_.MaximumIntensityProjectionZ(ext_nfo['lmp_soma'])
 
 # --- Extract all the extensions of all somas as a graph
 print('\n--- Graph extraction')
 
 # Create the graphs
 for soma in somas:
+    print(f" Soma {soma.uid}", end="")
     ext_map = skl_map_t.FromShapeMap(ext_nfo['lmp_soma'] == soma.uid, store_widths=True, skeletonize=False)  # do_post_thinning=True
     # to remove pixel that are not breaking connectivity - FineMap in FromShapeMap
     soma.skl_graph = skl_graph_t.FromSkeleton(ext_map, size_voxel=size_voxel_in_micron)
@@ -381,8 +393,6 @@ for soma in somas:
 
     # --- Find the root of the {ext+conn} graphs.
     # Roots are the nodes of degree 1 that are to be linked to the soma
-    print(f" Soma {soma.uid}", end="")
-
     soma.graph_roots = ge_.FindGraphsRootWithNodes(soma)
     # soma.graph_roots = ge_.FindGraphsRootWithEdges(soma, ext_nfo)
 
@@ -404,11 +414,16 @@ elapsed_time = tm_.gmtime(tm_.time() - start_time)
 # Extract features
 print('\n--- Features Extraction\n')
 # Prms
-number_of_bins = int(number_of_bins)
-max_range = hist_step * number_of_bins - hist_min_length
+number_of_bins_length = int(number_of_bins_length)
+bins_length = np_.linspace(hist_min_length, hist_min_length + hist_step_length * number_of_bins_length, num=number_of_bins_length)
+bins_length[-1] = np_.inf
+number_of_bins_curvature = int(number_of_bins_curvature)
+bins_curvature = np_.linspace(hist_min_curvature, hist_min_curvature + hist_step_curvature * number_of_bins_curvature, num=number_of_bins_curvature)
+bins_curvature[-1] = np_.inf
+
 # DF creation
-features_df = ge_.ExtractFeaturesInDF(somas, size_voxel_in_micron, number_of_bins, max_range, hist_min_length, ext_scales)
-# features_df.to_csv("...\features.csv")
+features_df = ge_.ExtractFeaturesInDF(somas, size_voxel_in_micron, bins_length, bins_curvature, ext_scales)
+features_df.to_csv("D:\\MorganeNadal\\M2 report\\features_512_.csv")
 
 #
 elapsed_time = tm_.gmtime(tm_.time() - start_time)

parameters.ini

@@ -54,7 +54,7 @@ channel                    :    G
 ; Can take the values R, G or B.
 ; Can also be set to None (cf. README above).
 size_voxel_in_micron
-; size_voxel_in_micron ->  X,Y,Z
+; size_voxel_in_micron ->  [X,Y,Z]
 ; Can also be set to None (cf. README above).
 save_images
 ; if None, no saving. Otherwise, path where to save images # TODO add in search_parameters.py
@@ -80,7 +80,7 @@ max_straight_sq_dist_c     :    0.24050024 * (30 ** 2)
 max_weighted_length_c      :    0.24050024 * 20.0
 
 [Frangi]
-scale_range                :    0.024,0.745
+scale_range                :    (0.024,0.745)
 ; Stop at max 3.1 (<=> 0.745) for number of Z stacks < 18
 ; write min,max scale range
 scale_step                 :    0.24
@@ -99,12 +99,18 @@ method                     :    c
 hist_min_length            :    0
 ; in micron
 ; by default, 0
-hist_step                  :    6
+hist_step_length           :    6
 ; in micron
-number_of_bins             :    5
+number_of_bins_length      :    5
 ; by default, 5
 ; extensions generally do not exceed 30 microns
 
+; histograms of extensions curvatures
+hist_min_curvature         :    0
+hist_step_curvature        :    6
+number_of_bins_curvature   :    5
+
+
 [Program Running]
 with_plot                  :    False
 in_parallel                :    False

search_parameters.py

@@ -38,25 +38,20 @@ parameters.read('parameters.ini')
 
 
 def IfNotFloat(section: str, key: str) -> float:
-    try:
-        value = parameters.getfloat(section, key)
-        return value
-    except:
-        return eval(parameters.get(section, key))
-
+    if parameters.get(section, key) is not None:
+        try:
+            value = parameters.getfloat(section, key)
+            return value
+        except:
+            return eval(parameters.get(section, key))
+    else:
+        return parameters.get(section, key)
 
 # [Input]
 data_path = parameters['Input']['data_path']
 channel = parameters['Input']['channel']
-
-try:
-    size_voxel_in_micron = parameters['Input']['size_voxel_in_micron']
-    if size_voxel_in_micron is not None:
-        size_voxel_in_micron = size_voxel_in_micron.split(',')
-        size_voxel_in_micron = list(float(size_voxel_in_micron[i]) for i in range(len(size_voxel_in_micron)))
-except:
-    raise ValueError('The voxel size parameter is not given correctly. Write X,Y,Z without () or [] or spaces. Do not '
-                     'forget ":" between size_voxel_in_micron and its value')
+size_voxel_in_micron = IfNotFloat('Input', 'size_voxel_in_micron')
+save_images = IfNotFloat('Input', 'save_images')
 
 # [Somas]
 soma_low_c = IfNotFloat('Somas', 'soma_low_c')
@@ -76,14 +71,7 @@ max_weighted_length_c = IfNotFloat('Connexions', 'max_weighted_length_c')
 
 # [Frangi]
 
-try:
-    scale_range = parameters['Frangi']['scale_range']
-    scale_range = scale_range.split(',')
-    scale_range = tuple(float(scale_range[i]) for i in range(len(scale_range)))
-except:
-    raise ValueError('The scale range parameter is not given correctly. Write min,max without () or [] or spaces. '
-                     'Do not forget ":" between scale_range and its value')
-
+scale_range = IfNotFloat('Frangi', 'scale_range')
 scale_step = IfNotFloat('Frangi', 'scale_step')
 alpha = IfNotFloat('Frangi', 'alpha')
 beta = IfNotFloat('Frangi', 'beta')
@@ -93,8 +81,11 @@ method = parameters['Frangi']['method']
 
 # [Features Extraction]
 hist_min_length = IfNotFloat('Features Extraction', 'hist_min_length')
-hist_step = IfNotFloat('Features Extraction', 'hist_step')
-number_of_bins = IfNotFloat('Features Extraction', 'number_of_bins')
+hist_step_length = IfNotFloat('Features Extraction', 'hist_step_length')
+number_of_bins_length = IfNotFloat('Features Extraction', 'number_of_bins_length')
+hist_min_curvature = IfNotFloat('Features Extraction', 'hist_min_curvature')
+hist_step_curvature = IfNotFloat('Features Extraction', 'hist_step_curvature')
+number_of_bins_curvature = IfNotFloat('Features Extraction', 'number_of_bins_curvature')
 
 # [Program running]
 with_plot = parameters.getboolean('Program Running', 'with_plot')

sklgraph/brick/edge.py

@@ -127,7 +127,6 @@ class edge_t:
         if check_validity:
             # A global condition: self.sites[0].size - 1 <= length
             if cast(array_t, sq_lengths == 0).any():
-                np_.save("D:\\MorganeNadal\\site_as_array_in_SetLength.npy", sites_as_array)
                 raise ValueError("Edge: Repeated sites")
             if cast(array_t, sq_lengths > self.sites.__len__()).any():
                 raise ValueError("Edge: Site gaps")