added test for model predict method

model.py

@@ -82,9 +82,6 @@ def _set_thresholds(self, train_data, generate_plots):
                 ]["player"].tolist()
 
                 number_of_flagged_players = len(all_flagged_players)
-                print(
-                    f"number of flagged players in {rating_bin_key} = {number_of_flagged_players}"
-                )
 
                 ## break if threshold is large enough to filter out all players
                 if number_of_flagged_players == 0:
@@ -94,9 +91,6 @@ def _set_thresholds(self, train_data, generate_plots):
                 for player in all_flagged_players:
                     self._player_account_handler.update_player_account_status(player)
 
-                print("updated account statuses:")
-                print(self._player_account_handler._account_statuses)
-
                 ## get the account status for each player
                 train_predictions = [
                     self._player_account_handler._account_statuses.get(player)

requirements.txt

@@ -8,5 +8,6 @@ python-lichess==0.10 # Client for lichess.org API
 mock==5.1.0
 pylint==3.0.3
 pytest==7.0.1
+pytest-dependency==0.6.0
 
 debugpy # Required for debugging.

tests/test_model.py

@@ -1,6 +1,7 @@
 import copy
 import unittest
 import pandas as pd
+from pandas.testing import assert_frame_equal
 import pytest
 from unittest import mock
 from model import PlayerAnomalyDetectionModel
@@ -38,7 +39,40 @@ def get_sample_train_data():
     return sample_train_data
 
 
+# fixture for sample test data
+# players 7-12 where players 10, 11, 12 perform above expected
+@pytest.fixture(scope="class")
+def get_sample_test_data():
+    sample_test_data = pd.DataFrame(
+        {
+            "player": [f"test_player{i}" for i in range(7, 6 + 7)] * 2,
+            "time_control": ["blitz"] * 6 + ["bullet"] * 6,
+            "number_of_games": [100] * 12,
+            "mean_perf_diff": [0.04, 0.06, 0.15, 0.161, 0.17, 0.25]
+            + [0.15, 0.15, 0.16, 0.171, 0.18, 0.25],
+            "std_perf_diff": [0.005] * 12,
+            "mean_rating": [1510, 1520, 1530, 1540, 1550, 1560]
+            + [1510, 1520, 1530, 1540, 1550, 1560],
+            "median_rating": [1510, 1520, 1530, 1540, 1550, 1560]
+            + [1510, 1520, 1530, 1540, 1550, 1560],
+            "std_rating": [10] * 12,
+            "mean_opponent_rating": [1510, 1520, 1530, 1540, 1550, 1560]
+            + [1510, 1520, 1530, 1540, 1550, 1560],
+            "std_opponent_rating": [10] * 12,
+            "mean_rating_gain": [1.00, -1.00, 1.00, -1.00, 1.00, -1.00]
+            + [1.00, -1.00, 1.00, -1.00, 1.00, -1.00],
+            "std_rating_gain": [0.01] * 12,
+            "proportion_increment_games": [1.00, 1.00, 1.00, 1.00, 1.00, 1.00]
+            + [0.00, 0.00, 0.00, 0.00, 0.00, 0.00],
+            "rating_bin": [1500] * 6 + [1600] * 6,
+        }
+    )
+
+    return sample_test_data
+
+
 @pytest.mark.usefixtures("get_sample_train_data", "build_training_data")
+@pytest.mark.usefixtures("get_sample_test_data", "build_training_data")
 class TestPlayerAnomalyDetectionModel(unittest.TestCase):
     @mock.patch(
         "player_account_handler.PlayerAccountHandler.update_player_account_status"
@@ -51,6 +85,10 @@ def setUp(self, mock_update_player_account_status):
     def build_training_data(self, get_sample_train_data):
         self.sample_train_data = get_sample_train_data
 
+    @pytest.fixture(autouse=True)
+    def build_test_data(self, get_sample_test_data):
+        self.sample_test_data = get_sample_test_data
+
     def test_fit(self):
         ## this is a workaround to avoid calling get_player_account_status
         self.model._player_account_handler._account_statuses = {
@@ -71,7 +109,46 @@ def test_fit(self):
         assert expected_thresholds == self.model._thresholds
 
     def test_predict(self):
-        pass
+        self.model._player_account_handler._account_statuses = {
+            "test_player7": "open",
+            "test_player8": "open",
+            "test_player9": "open",
+            "test_player10": "closed",
+            "test_player11": "tosViolation",
+            "test_player12": "tosViolation",
+        }
+        self.model.fit(self.sample_test_data, generate_plots=False)
+        expected_predictions = self.sample_test_data.copy()
+        expected_predictions["is_anomaly"] = [
+            False,
+            False,
+            False,
+            True,
+            True,
+            True,
+            False,
+            False,
+            False,
+            True,
+            True,
+            True,
+        ]
+        expected_predictions["account_status"] = [
+            "open",
+            "open",
+            "open",
+            "closed",
+            "tosViolation",
+            "tosViolation",
+            "open",
+            "open",
+            "open",
+            "closed",
+            "tosViolation",
+            "tosViolation",
+        ]
+        test_predictions = self.model.predict(self.sample_test_data)
+        assert_frame_equal(expected_predictions, test_predictions)
 
     def test_save_model(self):
         pass