Added recommendations 'Algorithm'

Author: Seluj78

backend/PyMatcha/routes/api/recommendations.py

@@ -0,0 +1,35 @@
+"""
+    PyMatcha - A Python Dating Website
+    Copyright (C) 2018-2019 jlasne/gmorer
+    <jlasne@student.42.fr> - <gmorer@student.42.fr>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+"""
+from flask import Blueprint
+from flask_jwt_extended import current_user
+from flask_jwt_extended import jwt_required
+from PyMatcha import redis
+from PyMatcha.utils.errors import NotFoundError
+from PyMatcha.utils.success import SuccessOutput
+
+recommendations_bp = Blueprint("recommendations", __name__)
+
+
+@recommendations_bp.route("/recommendations", methods=["GET"])
+@jwt_required
+def get_recommendations():
+    recommendations = redis.get(f"user_recommendations:{str(current_user.id)}")
+    if not recommendations:
+        raise NotFoundError("Recommendations not calculated yet", "Please come back later")
+    return SuccessOutput("recommendations", recommendations)

backend/PyMatcha/utils/match_score.py

@@ -0,0 +1,36 @@
+from typing import List
+
+from fuzzywuzzy import fuzz
+from Geohash import decode
+from geopy.distance import distance
+
+
+def _get_distance(geohash_1: str, geohash_2: str) -> float:
+    coords_1 = decode(geohash_1)
+    coords_2 = decode(geohash_2)
+    return distance(coords_1, coords_2).kilometers
+
+
+def _get_common_tags(tag_list_1: list, tag_list_2: list) -> List[str]:
+    """Will return the list of common tags
+    """
+    common_tags = []
+    for tag_1 in tag_list_1:
+        for tag_2 in tag_list_2:
+            if fuzz.partial_ratio(tag_1, tag_2) >= 70:
+                common_tags.append(tag_1 if len(tag_1) > len(tag_2) else tag_2)
+    return list(set(common_tags))
+
+
+def _get_age_diff(age_1: int, age_2: int) -> int:
+    return -1 * (age_1 - age_2)
+
+
+def _get_inverted_gender(gender, orientation):
+    # TODO: Handle other gender and bisexual and other orientations
+    if orientation == "heterosexual":
+        return "male" if gender == "female" else "female"
+    elif orientation == "homosexual":
+        return "male" if gender == "male" else "female"
+    else:
+        return "other"

backend/PyMatcha/utils/tasks.py

@@ -17,6 +17,7 @@
 def setup_periodic_tasks(sender, **kwargs):
     sender.add_periodic_task(60, update_offline_users.s(), name="Update online users every minute")
     sender.add_periodic_task(3600, update_heat_scores.s(), name="Update heat scores every hour")
+    sender.add_periodic_task(60, update_user_recommendations.s(), name="Update user recommendations every minute")
 
 
 @celery.task
@@ -92,3 +93,60 @@ def update_offline_users():
     return "Updated online status for {} users. {} passed offline and {} passed or stayed online.".format(
         count, offline_count, online_count
     )
+
+
+def default_date_converter(o):
+    if isinstance(o, datetime.datetime):
+        return o.__str__()
+
+
+@celery.task
+def update_user_recommendations():
+    today = datetime.datetime.utcnow()
+    count = 0
+    for user_to_update in User.select_all():
+        count += 1
+        user_to_update_recommendations = []
+
+        user_to_update_age = (
+            today.year
+            - user_to_update.birthdate.year
+            - ((today.month, today.day) < (user_to_update.birthdate.month, user_to_update.birthdate.day))
+        )
+        user_to_update_tags = [t.name for t in user_to_update.get_tags()]
+
+        inverted_gender = _get_inverted_gender(user_to_update.gender, user_to_update.orientation)
+
+        for user in User.get_multis(orientation=user_to_update.orientation, gender=inverted_gender):
+            if user.id == user_to_update.id:
+                continue
+            score = 0
+
+            distance = _get_distance(user_to_update.geohash, user.geohash)
+            score -= distance
+
+            user_age = (
+                today.year
+                - user.birthdate.year
+                - ((today.month, today.day) < (user.birthdate.month, user.birthdate.day))
+            )
+            age_diff = _get_age_diff(user_to_update_age, user_age)
+            score -= age_diff
+
+            user_tags = [t.name for t in user.get_tags()]
+            common_tags = _get_common_tags(user_to_update_tags, user_tags)
+            score += len(common_tags) * 2
+
+            score += user.heat_score
+
+            d = {"score": score, "common_tags": common_tags, "distance": distance}
+            d.update(user.to_dict())
+            user_to_update_recommendations.append(d)
+        user_to_update_recommendations_sorted = sorted(
+            user_to_update_recommendations, key=lambda x: x["score"], reverse=True
+        )
+        redis.set(
+            f"user_recommendations:{str(user_to_update.id)}",
+            json.dumps(user_to_update_recommendations_sorted, default=default_date_converter),
+        )
+    return f"Successfully updated recommendations for {count} users."

backend/requirements.txt

@@ -29,4 +29,8 @@ lorem==0.1.1
 future==0.18.2
 ip2geotools==0.1.5
 
-argon2-cffi==20.1.0
+argon2-cffi==20.1.0
+
+geopy==2.0.0
+fuzzywuzzy==0.18.0
+python-Levenshtein==0.12.0