From d2ab3264ca3e432db44c43a34f0641d90f465c0b Mon Sep 17 00:00:00 2001
From: Joseph Kuo <chao-chung.kuo@rwth-aachen.de>
Date: Tue, 20 Feb 2024 15:14:18 +0100
Subject: [PATCH] remove intersect_posi

---
 genomkit/regions/gregions.py         | 43 ----------------------------
 tests/evaluate/evaluate_intersect.py |  6 ----
 2 files changed, 49 deletions(-)

diff --git a/genomkit/regions/gregions.py b/genomkit/regions/gregions.py
index c700a53..d313946 100644
--- a/genomkit/regions/gregions.py
+++ b/genomkit/regions/gregions.py
@@ -598,49 +598,6 @@ def make_array_same_length(array_1, array_2):
                                 name=""))
         return res
 
-    def intersect_posi(self, target, strandness: bool = False):
-        def generate_position_array(position_set, merged_list):
-            boolean_array = [False] * len(merged_list)
-            # Set elements to True according to pos_1
-            for i, element in enumerate(merged_list):
-                if element in position_set:
-                    boolean_array[i] = True
-            return np.array(boolean_array)
-
-        def find_intersects(orientation):
-            pos_1 = self.get_positions_by_seq(sequence=seq,
-                                              orientation=orientation)
-            pos_2 = target.get_positions_by_seq(sequence=seq,
-                                                orientation=orientation)
-            merged_pos = list(pos_1.union(pos_2))
-            array_1 = generate_position_array(pos_1, merged_pos)
-            array_2 = generate_position_array(pos_2, merged_pos)
-            result_array = array_1 & array_2
-            # Find the indices where the value changes
-            indices = np.where(np.diff(result_array))[0] + 1
-            # Group consecutive indices into tuples
-            ranges = [(merged_pos[indices[i - 1]],
-                       merged_pos[indices[i]])
-                      for i in range(1, len(indices), 2)]
-            return ranges
-
-        res = GRegions()
-        list_seq_self = self.get_sequences(unique=True)
-        list_seq_target = target.get_sequences(unique=True)
-        common_seq = [seq for seq in list_seq_self if seq in list_seq_target]
-        for seq in common_seq:
-            if strandness:
-                # positive
-                ranges_pos = find_intersects(orientation="+")
-                ranges_neg = find_intersects(orientation="-")
-                ranges = ranges_pos + ranges_neg
-            else:
-                ranges = find_intersects(orientation=None)
-            for pair in ranges:
-                res.add(GRegion(sequence=seq, start=pair[0], end=pair[1],
-                                name=""))
-        return res
-
     def overlap_count(self, target):
         intersect = self.intersect_python(target, mode="ORIGINAL")
         return len(intersect)
diff --git a/tests/evaluate/evaluate_intersect.py b/tests/evaluate/evaluate_intersect.py
index 434ca3c..5b33422 100644
--- a/tests/evaluate/evaluate_intersect.py
+++ b/tests/evaluate/evaluate_intersect.py
@@ -30,17 +30,11 @@ def time_intersect_python():
 def time_intersect_array():
     intersect = peaks.intersect_array(genes)
 
-@profile
-def time_intersect_posi():
-    intersect = peaks.intersect_posi(genes)
-
 repeat_num = 2
 execution_time = timeit.timeit(time_intersect_python, number=repeat_num)
 print('[{:<20}]'.format('intersect_python'), '{:<5.2f}'.format(execution_time), "seconds")
 execution_time = timeit.timeit(time_intersect_array, number=repeat_num)
 print('[{:<20}]'.format('intersect_array'), '{:<5.2f}'.format(execution_time), "seconds")
-execution_time = timeit.timeit(time_intersect_posi, number=repeat_num)
-print('[{:<20}]'.format('intersect_posi'), '{:<5.2f}'.format(execution_time), "seconds")
 # print(len(intersect))
 # intersect.write("intersect_python.bed")
 # print(len(intersect))