Fixed inconsistency with start_end_array_to_bool_array and fixed tests

CHANGELOG.md

@@ -80,6 +80,11 @@ project.
   Both wrapper (Stride-level and Region) are both aware of this parameter and pass them correctly down to the respective
   methods.
   (https://github.com/mad-lab-fau/gaitmap/pull/14)
+- The util method `start_end_array_to_bool_array` now assumes that the end index of all regions is inclusive.
+  This enables roundtrip conversion with the `bool_array_to_start_end_array` method and is in line with the definitions
+  used for strides, ROIs, and ZUPTs in gaitmap.
+  (https://github.com/mad-lab-fau/gaitmap/pull/14)
+
 
 ### Removed
 - The `find_zupts` method of the `RtsKalmanFilter` and all deprecated arguments are now fully removed in favor of the 

gaitmap/utils/array_handling.py

@@ -141,6 +141,9 @@ def start_end_array_to_bool_array(start_end_array: np.ndarray, pad_to_length: in
     start_end_array : array with shape (n,2)
         2d-array indicating start and end values e.g. [[10,20],[20,40],[70,80]]
 
+        .. warning:: We assume that the end index is exclusiv!
+                     This is in line with the definitions of stride and roi lists in gaitmap.
+
     pad_to_length: int
         define length of resulting array if None is given the array will have the length of the last element of the
         initial start_end_array
@@ -152,24 +155,28 @@ def start_end_array_to_bool_array(start_end_array: np.ndarray, pad_to_length: in
 
     Examples
     --------
-    >>> example_array = np.array([[3,5],[7,8], pad_to_length=12)
-    >>> bool_array = start_end_array_to_bool_array(example_array)
-    >>> bool_array
-    array([False,False,False,True,True,True,False,True,True, False, False, False])
-    >>>  example_array = np.array([[3,5],[7,8], pad_to_length=None)
-    array([False,False,False,True,True,True,False,True,True])
+    >>> import numpy as np
+    >>> example_array = np.array([[3,5],[7,8]])
+    >>> start_end_array_to_bool_array(example_array, pad_to_length=12)
+    array([False, False, False,  True,  True,  True, False,  True,  True,
+           False, False, False])
+    >>> example_array = np.array([[3,5],[7,8]])
+    >>> start_end_array_to_bool_array(example_array, pad_to_length=None)
+    array([False, False, False,  True,  True,  True, False,  True,  True])
 
     """
     start_end_array = np.atleast_2d(start_end_array)
-    if pad_to_length and pad_to_length <= start_end_array[-1][-1]:
-        raise ValueError("Padding length must be larger than last element of start end array!")
 
-    n_elements = start_end_array[-1][-1] + 1
+    n_elements = start_end_array.max()
+
     if pad_to_length:
+        if pad_to_length <= n_elements:
+            raise ValueError("Padding length must be larger than last element of start end array!")
         n_elements = pad_to_length
+
     bool_array = np.zeros(n_elements)
     for start, end in start_end_array:
-        bool_array[start : end + 1] = 1
+        bool_array[start:end] = 1
     return bool_array.astype(bool)
 
 

tests/test_utils/test_array_handling.py

@@ -254,25 +254,34 @@ class TestStartEndArrayToBoolArray:
     def test_simple_input_no_padding(self):
         input_array = np.array([[2, 3], [5, 9]])
         output_array = start_end_array_to_bool_array(input_array)
-        expected_output = np.array([0, 0, 1, 1, 0, 1, 1, 1, 1, 1]).astype(bool)
+        expected_output = np.array([0, 0, 1, 0, 0, 1, 1, 1, 1]).astype(bool)
         assert_array_equal(expected_output, output_array)
 
     def test_simple_input_1d_no_padding(self):
         input_array = np.array([2, 3])
         output_array = start_end_array_to_bool_array(input_array, pad_to_length=5)
-        expected_output = np.array([0, 0, 1, 1, 0]).astype(bool)
+        expected_output = np.array([0, 0, 1, 0, 0]).astype(bool)
         assert_array_equal(expected_output, output_array)
 
     def test_simple_input_with_padding(self):
         input_array = np.array([[2, 3], [5, 9]])
         output_array = start_end_array_to_bool_array(input_array, pad_to_length=12)
-        expected_output = np.array([0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0]).astype(bool)
+        expected_output = np.array([0, 0, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0]).astype(bool)
+        assert_array_equal(expected_output, output_array)
+
+    @pytest.mark.parametrize("reverse", [True, False])
+    def test_length_unsorted(self, reverse):
+        input_array = np.array([[5, 9], [2, 3]])
+        if reverse:
+            input_array = input_array[::-1]
+        output_array = start_end_array_to_bool_array(input_array)
+        expected_output = np.array([0, 0, 1, 0, 0, 1, 1, 1, 1]).astype(bool)
         assert_array_equal(expected_output, output_array)
 
     def test_overlapping_input_with_padding(self):
         input_array = np.array([[2, 6], [5, 9]])
         output_array = start_end_array_to_bool_array(input_array, pad_to_length=12)
-        expected_output = np.array([0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0]).astype(bool)
+        expected_output = np.array([0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0]).astype(bool)
         assert_array_equal(expected_output, output_array)
 
     def test_invalid_padding(self):
@@ -287,6 +296,12 @@ def test_correct_output_dtype(self):
         output_array = start_end_array_to_bool_array(input_array)
         assert output_array.dtype == "bool"
 
+    def test_bool_array_round_trip(self):
+        input_array = np.array([[2, 3], [5, 9]])
+        output_array = start_end_array_to_bool_array(input_array)
+        output_array = bool_array_to_start_end_array(output_array)
+        assert_array_equal(input_array, output_array)
+
 
 class TestLocalMinimaBelowThreshold:
     @pytest.mark.parametrize(