Adding Longest Common Subsequence

pydatastructs/linear_data_structures/__init__.py

@@ -30,6 +30,7 @@
     counting_sort,
     bucket_sort,
     cocktail_shaker_sort,
-    quick_sort
+    quick_sort,
+    longest_common_subsequence
 )
 __all__.extend(algorithms.__all__)

pydatastructs/linear_data_structures/algorithms.py

@@ -13,7 +13,8 @@
     'counting_sort',
     'bucket_sort',
     'cocktail_shaker_sort',
-    'quick_sort'
+    'quick_sort',
+    'longest_common_subsequence'
 ]
 
 def _merge(array, sl, el, sr, er, end, comp):
@@ -722,3 +723,64 @@ def partition(low, high, pick_pivot_element):
         array._modify(force=True)
 
     return array
+
+def longest_common_subsequence(seq1, seq2) -> tuple:
+    """
+    Implements Longest Common Subsequence
+
+    Parameters
+    ========
+
+    seq1: Any 1D data structure that can be indexed (like list, tuple, string)
+    seq2: Any 1D data structure that can be indexed (like list, tuple, string)
+
+    Returns
+    =======
+
+    output: tuple
+        The first element of the tuple represents the length of longest common subsequence and
+        the second element is the longest common subsequence itself.
+        Common subsequence will be of the same data type as that of input sequences.
+
+    Examples
+    ========
+
+    >>> from pydatastructs import longest_common_subsequence as LCS
+    >>> LCS("ABCDEF", "ABBCDDDE")
+    (5, 'ABCDE')
+    >>> arr1 = ['A', 'P', 'P']
+    >>> arr2 = ['A', 'p', 'P', 'S', 'P']
+    >>> LCS(arr1, arr2)
+    (3, ['A', 'P', 'P'])
+
+    References
+    ==========
+
+    .. [1] https://en.wikipedia.org/wiki/Longest_common_subsequence_problem
+    """
+    row, col = len(seq1), len(seq2)
+    check_mat = {0:[0 for _ in range(col+1)]}
+    for i in range(row):
+        check_mat[i+1]=[0 for _ in range(col+1)]
+        for j in range(col):
+            if (seq1[i] == seq2[j]):
+                check_mat[i+1][j+1] = check_mat[i][j]+1
+            else:
+                check_mat[i+1][j+1] = max(check_mat[i+1][j], check_mat[i][j+1])
+
+    lcseq, lclen = [], check_mat[row][col]
+    while(row > 0  and col > 0):
+        if(check_mat[row][col] == check_mat[row][col-1]):
+            col -= 1
+        elif(check_mat[row][col] == check_mat[row-1][col]):
+            row -= 1
+        else:
+            lcseq.append(seq1[row-1])
+            row -= 1
+            col -= 1
+
+    if(type(seq1) == str):
+        lcseq = ''.join(lcseq)
+    if(type(seq1) == tuple):
+        lcseq = tuple(lcseq)
+    return (lclen, lcseq[::-1])

pydatastructs/linear_data_structures/tests/test_algorithms.py

@@ -1,7 +1,8 @@
 from pydatastructs import (
     merge_sort_parallel, DynamicOneDimensionalArray,
     OneDimensionalArray, brick_sort, brick_sort_parallel,
-    heapsort, matrix_multiply_parallel, counting_sort, bucket_sort, cocktail_shaker_sort, quick_sort)
+    heapsort, matrix_multiply_parallel, counting_sort, bucket_sort, cocktail_shaker_sort, quick_sort, longest_common_subsequence)
+
 
 from pydatastructs.utils.raises_util import raises
 import random
@@ -100,3 +101,22 @@ def test_matrix_multiply_parallel():
     J = [[2, 1, 2], [1, 2, 1], [2, 2, 2]]
     output = matrix_multiply_parallel(I, J, num_threads=1)
     assert expected_result == output
+
+def test_longest_common_sequence():
+    expected_result = (5, 'ASCII')
+
+    str1, str2 = 'AASCCII', 'ASSCIIII'
+    output = longest_common_subsequence(str1, str2)
+    assert expected_result == output
+
+    expected_result = (3, ['O', 'V', 'A'])
+
+    I, J = ['O', 'V', 'A', 'L'], ['F', 'O', 'R', 'V', 'A', 'E', 'W']
+    output = longest_common_subsequence(I, J)
+    assert expected_result == output
+
+    expected_result = (2, ('O', 't'))
+
+    I, J = ('K', 'O', 't', 'H', 'E'), ('L', 'O', 'C', 't')
+    output = longest_common_subsequence(I, J)
+    assert expected_result == output