Added Gale-Shapley algorithm with test cases (#1963)

algorithms/greedy/__init__.py

@@ -1 +1,2 @@
 from .max_contiguous_subsequence_sum import *
+from .gale_shapley import *

algorithms/greedy/gale_shapley.py

@@ -0,0 +1,90 @@
+"""
+The Gale-Shapley algorithm is a method to solve the
+stable matching/marriage problem. Given N men and women
+with ranked preferences of the opposite sex, the men and
+women will be matched so that each pair of man and woman
+would not prefer someone over their current match. With
+no conflicting preferences, the matches are stable. The
+algorithm can be extended to issues that involve ranked
+matching.
+
+Example for function:
+    M denotes man, W denotes woman, and number corresponds to
+    respective man/woman. Preference lists go from highest to lowest.
+
+    men_preferences = {
+        "M1": ["W1", "W2", "W3"],
+        "M2": ["W1", "W3", "W2"],
+        "M3": ["W3", "W1", "W2"],
+    }
+    women_preferences = {
+        "W1": ["M2", "M1", "M3"],
+        "W2": ["M1", "M2", "M3"],
+        "W3": ["M3", "M1", "M2"],
+    }
+
+    input: print(gale_shapley(men_preferences, women_preferences))
+    output: {'M2': 'W1', 'M3': 'W3', 'M1': 'W2'}
+    Explanation:
+        Both Man 1 and Man 2 have a top preference of Woman 1,
+        and since Woman 1 has a top preference of Man 2, Man 2
+        and Woman 1 are matched, and Man 1 is added back to available
+        men. Man 3 and Woman 3 have their top preference as each other,
+        so the two are matched. Man 1 then proposes to Woman 2, and
+        Man 1 is the top preference of Woman 2, so the two are matched.
+        There is no match of Man AND Woman where both would want to
+        leave, so the current matches are stable.
+
+"""
+
+# size denotes the number of men/women
+# Function takes in dictionary for men and women preferences in style outlined above
+def gale_shapley(men, women):
+    size = len(men)
+    # Initialize all men to be available
+    men_available = list(men.keys())
+    # Initialize married to empty
+    married = {}
+    # Intialize proposal count for each man to 0
+    proposal_counts = {man: 0 for man in men}
+    while men_available:
+        # Pop first available man
+        man = men_available.pop(0)
+        # Of the popped man, set woman equal to corresponding proposal index
+        woman = men[man][proposal_counts[man]]
+        #increment proposal count
+        proposal_counts[man] += 1
+        if woman not in married:
+            # Set marriage if woman not married
+            married[woman] = man
+        else:
+            # If woman married, currently_married corresponds to currently matched man
+            currently_married = married[woman]
+            if women[woman].index(man) < women[woman].index(currently_married):
+                """
+                If the available man is of greater preference to the woman than her
+                currently married partner, change the marriage to the new available
+                man and append the previously married man back to men_available
+                """
+                married[woman] = man
+                men_available.append(currently_married)
+            else:
+                # Add man back to men_available and try woman at next index
+                men_available.append(man)
+    # Returns pairs of matched men and women in form of Man:Woman
+    return {man: woman for woman, man in married.items()}
+
+# Example case
+men_preferences = {
+    "M1": ["W1", "W2", "W3"],
+    "M2": ["W1", "W3", "W2"],
+    "M3": ["W3", "W1", "W2"],
+}
+women_preferences = {
+    "W1": ["M2", "M1", "M3"],
+    "W2": ["M1", "M2", "M3"],
+    "W3": ["M3", "M1", "M2"],
+}
+
+res = gale_shapley(men_preferences, women_preferences)
+print(res)