Applies suggestions.

pythia/seo/elo.py

@@ -1,5 +1,6 @@
 import pandas as pd
 import numpy as np
+from collections import deque
 
 
 class ELO:
@@ -103,9 +104,9 @@ def score_update(self, image_0, image_1, score_for_image_0):
         """
         # state dicts
         state_dict_0 = self.rankings.loc[image_0].to_dict()
-        state_dict_0['last_scores'] = list(map(float, state_dict_0['last_scores'].split(',')))
+        state_dict_0['last_scores'] = deque(map(float, state_dict_0['last_scores'].split(',')), maxlen=self.score_memory)
         state_dict_1 = self.rankings.loc[image_1].to_dict()
-        state_dict_1['last_scores'] = list(map(float, state_dict_1['last_scores'].split(',')))
+        state_dict_1['last_scores'] = deque(map(float, state_dict_1['last_scores'].split(',')), maxlen=self.score_memory)
 
         expected_score_0 = self.expected_score(self.rankings.loc[image_0]['score'],
                                                self.rankings.loc[image_1]['score'])
@@ -119,15 +120,10 @@ def score_update(self, image_0, image_1, score_for_image_0):
         state_dict_0['last_scores'].append(new_rating_0)
         state_dict_1['last_scores'].append(new_rating_1)
 
-        if len(state_dict_0['last_scores']) > self.score_memory:
-            state_dict_0['last_scores'].pop(0)
-        if len(state_dict_1['last_scores']) > self.score_memory:
-            state_dict_1['last_scores'].pop(0)
-
-        new_std_dev_0 = min(np.std(state_dict_0['last_scores']), 1000000)  # prevents Infinity
+        new_std_dev_0 = min(np.std(state_dict_0['last_scores']), 1_000_000)  # prevents Infinity
         new_k_0 = min(max(new_std_dev_0, self.score_change['min']), self.score_change['max'])
 
-        new_std_dev_1 = min(np.std(state_dict_1['last_scores']), 1000000)  # prevents Infinity
+        new_std_dev_1 = min(np.std(state_dict_1['last_scores']), 1_000_000)  # prevents Infinity
         new_k_1 = min(max(new_std_dev_1, self.score_change['min']), self.score_change['max'])
 
         # Updating Data