Refactoring test script and updated requirements

environment.yml

@@ -4,41 +4,38 @@ channels:
   - conda-forge
   - defaults
 dependencies:
-  - matplotlib==3.7.0
-  - nibabel==3.2.2
-  - pandas==1.5.3
-  - pillow==9.2.0
-  - pydicom=1.2.2
-  - pywavelets==1.1.1
-  - scikit-image==0.18.2
-  - scipy==1.7.0
-  - setuptools==62.2.0
-  - tqdm==4.65.0
+  - matplotlib
+  - nibabel
+  - pandas<2.0.0
+  - pillow
+  - pydicom
+  - pywavelets
+  - scikit-image
+  - scipy
+  - setuptools
+  - tqdm
   - pip
   - pip:
-    - numpy==1.22.4
-    - isort==5.10.1
-    - ipykernel==6.15.1
-    - ipywidgets==8.0.1
-    - jupyter==1.0.0
-    - networkx==2.8.5
-    - neuroCombat==0.2.12
-    - nilearn==0.10.1
-    - numpyencoder==0.3.0
-    - protobuf==3.20.*
-    - pycaret==3.0.4
-    - ray==2.5.1
-    - scikit_image==0.18.2
-    - SimpleITK==2.1.1.2
-    - scikit_learn==1.2.2
-    - seaborn==0.13.2
-    - Sphinx==7.2.6
+    - numpy
+    - isort
+    - ipykernel
+    - ipywidgets
+    - jupyter
+    - networkx
+    - neuroCombat
+    - nilearn
+    - numpyencoder
+    - protobuf
+    - pycaret
+    - ray
+    - scikit_image
+    - SimpleITK
+    - scikit_learn
+    - seaborn
+    - Sphinx
     - sphinx-carousel==1.2.0
-    - sphinx-autodoc-typehints==1.19.2
-    - sphinx-jinja2-compat==0.1.2
     - sphinx-jsonschema==1.19.1
     - sphinx-rtd-dark-mode==1.2.4
-    - sphinx-rtd-theme==0.5.2
-    - wget==3.2
-    - tabulate==0.9.0
-    - xgboost==1.7.6
+    - wget
+    - tabulate
+    - xgboost

requirements.txt

@@ -9,7 +9,7 @@ neuroCombat
 nibabel
 nilearn
 numpyencoder
-pandas
+pandas<2.0.0
 Pillow
 protobuf
 pycaret

tests/test_extraction.py

@@ -10,43 +10,105 @@
 
 
 class TestExtraction:
-
-    def _get_phantom(self):
-        phantom = np.zeros((64,64,64))
-        phantom[32,32,32] = 255
+    """
+    Test the extraction of morphological and statistical features.
+
+    Features are extracted from the IBSI phantom, and values are compared to the IBSI ones.
+
+    Phantom and reference values can be found in the IBSI manual: https://arxiv.org/pdf/1612.07003.pdf
+    """
+    def __get_phantom(self):
+        phantom = np.array(
+            [[
+                [1, 1, 1, 1,],
+                [1, 1, 1, 1,],
+                [4, 1, 1, 1,],
+                [4, 4, 1, 1,]],
+
+                [[4, 4, 4, 4,],
+                [4, 1, 1, 1,],
+                [1, 1, 1, 1,],
+                [4, 4, 1, 1,]],
+
+                [[4, 4, 4, 4,],
+                [6, 6, 1, 1,],
+                [6, 3, 9, 1,],
+                [6, 6, 6, 6,]],
+
+                [[1, 1, 1, 1,],
+                [1, 1, 1, 1,],
+                [4, 1, 1, 1,],
+                [4, 1, 1, 1,]],
+
+                [[1, 1, 1, 1,],
+                [1, 1, 1, 1,],
+                [1, 1, 1, 1,],
+                [1, 1, 1, 1,]]], dtype=np.float32
+        )
 
         return phantom
 
-    def _get_random_roi(self):
-        roi = np.zeros((64,64,64))
-        roi[
-            np.random.randint(0,64,5),
-            np.random.randint(0,64,5),
-            np.random.randint(0,64,5)] = 1
+    def __get_random_roi(self):
+        roi = np.array(
+            [[
+                [1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 0, 1, 1],
+                [1, 1, 1, 1]],
+
+                [[1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1]],
+
+                [[1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 1, 0, 1],
+                [1, 1, 1, 1]],
+
+                [[1, 1, 0, 0],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1]],
+
+                [[1, 1, 0, 0],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1],
+                [1, 1, 1, 1]]], dtype=np.int16
+        )
+
         return roi
 
     def test_morph_features(self):
-        phantom = self._get_phantom()
-        roi = self._get_random_roi()
+        phantom = self.__get_phantom()
+        roi = self.__get_random_roi()
         morph = MEDimage.biomarkers.morph.extract_all(
             vol=phantom, 
             mask_int=roi, 
             mask_morph=roi,
-            res=[1,1,1],
+            res=[2, 2, 2],
             intensity_type="arbitrary"
         )
         morph_vol = MEDimage.biomarkers.morph.vol(
             vol=phantom, 
             mask_int=roi, 
             mask_morph=roi,
-            res=[1,1,1]
+            res=[2, 2, 2]
+        )
+        surface_area = MEDimage.biomarkers.morph.area(
+            vol=phantom, 
+            mask_int=roi, 
+            mask_morph=roi,
+            res=[2, 2, 2]
         )
         assert morph_vol == morph["Fmorph_vol"]
-        assert round(morph_vol, 2) == 0.83
+        assert abs(morph_vol - 556) < 1
+        assert surface_area == morph["Fmorph_area"]
+        assert abs(surface_area - 388) < 1
 
     def test_stats_features(self):
-        phantom = self._get_phantom()
-        roi = self._get_random_roi()
+        phantom = self.__get_phantom()
+        roi = self.__get_random_roi()
         vol_int_re = MEDimage.processing.roi_extract(
             vol=phantom, 
             roi=roi
@@ -58,5 +120,10 @@ def test_stats_features(self):
         kurt = MEDimage.biomarkers.stats.kurt(
             vol=vol_int_re,
         )
+        skewness = MEDimage.biomarkers.stats.skewness(
+            vol=vol_int_re,
+        )
         assert kurt == stats["Fstat_kurt"]
-        assert round(kurt, 2) == -3.0
+        assert abs(kurt + 0.355) < 0.01
+        assert skewness == stats["Fstat_skew"]
+        assert abs(skewness - 1.08) < 0.01