Usuń 'df50.csv'

Reviewed-on: #1

data_dictionary.txt

@@ -1,4 +1,4 @@
-link_match		: page link of match in Premier League Official website
+link_match		: page link of match in Premier League Official websitemarkdow
 season			: match season
 date			: match date
 home_team		: home team

data_filters.py

@@ -1,5 +1,17 @@
 import pandas as pd
-from simpful import *
+from fuzzy import *
+
+
+def save_to_csv(filename, dataframe):
+    dataframe.to_csv(filename, mode='a', index=False, header=not pd.DataFrame().append(dataframe).empty)
+
+def split_to_parts(dataframe, part_size):
+    for i in range(0, len(dataframe), part_size):
+        yield dataframe.iloc[i:i + part_size]
+
+def przetwarzaj_co_50_rekordow(plik_wejsciowy, plik_wyjsciowy):
+    dataframe_wejsciowe = pd.read_csv(plik_wejsciowy)
+
 
 def generateTrainingData(dataframe):
     columns = ['season','date','home_team','away_team','result_full','home_passes','away_passes',
@@ -10,25 +22,45 @@ def generateTrainingData(dataframe):
 def generateFuzzyLogicData(dataframe):
     columns = ['season','date','home_team','away_team','result_full','c_home_form_5m','c_away_form_5m',#,'c_home_passes','c_away_passes',
               # 'c_home_possession','c_away_possession','c_home_shots','c_away_shots',
-               'c_home_diff_5m', 'c_away_diff_5m','c_home_form_season','c_away_form_season',
+               'c_home_diff_5m', 'c_away_diff_5m',"c_home_form_5s",
-               'c_home_diff_season', 'c_away_diff_season']
+               'c_away_form_5s','c_home_diff_5s','c_away_diff_5s'
+               , 'c_home_aggression_5m',
+               'c_away_aggression_5m', 'c_away_shots_5m','c_away_shots_5m',
+               'c_away_shots_5btw', 'c_away_shots_5btw', 'c_away_defence_5m',
+               'c_away_defence_5m', 'c_away_defence_5btw', 'c_away_defence_5btw',
+               'c_home_passing_5m', 'c_away_passing_5m', 'c_home_passing_5btw',
+               'c_away_passing_5btw', 'c_away_aggression_5btw', 'c_away_aggression_5btw'
+
+
+               #'c_home_aggression_season', 'c_away_aggression_season',
+             #  'c_home_form_season','c_away_form_season',
+            #   'c_home_diff_season', 'c_away_diff_season'
+               ]
     return dataframe[columns]
 
+
 def last5Matches(season, teamA, data, df):
-    # Wybierz rekordy dla danej pary drużyn i sezonu
+    
     subset = df[((df['season'] == season) & ((df['home_team'] == teamA) | (df['away_team'] == teamA)))]
 
-    # Filtruj dane, aby zawierały te przed daną datą
+    
     before_given_date = subset[pd.to_datetime(subset['date']) < pd.to_datetime(data)]
 
-    # Posortuj wg daty w odwrotnej kolejności
+   
     before_given_date = before_given_date.sort_values(by='date', ascending=False)
 
-    # Wybierz 5 ostatnich przed daną datą
+    
     last_before_date = before_given_date.head(5)
 
-    return last_before_date
+    return last_before_date, "_5m"
 
+def last5MatchesBtwTeams(teamA, teamB, data, df):
+    subset = df[(((df['home_team'] == teamA) | (df['away_team'] == teamA)) & ((df['home_team'] == teamB) | (df['away_team'] == teamB)))]
+    before_given_date = subset[pd.to_datetime(subset['date']) < pd.to_datetime(data)]
+    before_given_date = before_given_date.sort_values(by='date', ascending=False)
+    last_before_date = before_given_date.head(5)
+
+    return last_before_date, "_5btw"
 
 def seasonMatches(season, teamA, data, df):
     # Wybierz rekordy dla danej pary drużyn i sezonu
@@ -40,7 +72,7 @@ def seasonMatches(season, teamA, data, df):
     # Posortuj wg daty w odwrotnej kolejności
     before_given_date = before_given_date.sort_values(by='date', ascending=False)
 
-    return before_given_date
+    return before_given_date, "_s"
 
 def getResult(score,teamHome):
     x,y = score.split('-')
@@ -53,7 +85,24 @@ def getResult(score,teamHome):
         return "draw"
     else:
         return "loss"
-    
+
+# def calculateAggression(matches, team):
+#     aggression = 0
+#     for index, row in matches.iterrows():
+#         if team == row['home_team']:
+#             yellow_cards = row['home_yellow_cards']
+#             red_cards = row['home_red_cards']
+#         else:
+#             yellow_cards = row['away_yellow_cards']
+#             red_cards = row['away_red_cards']
+#         aggression_result = calculateFuzzyAggression(yellow_cards, red_cards)
+#         #print(aggression_result['aggression'])
+#         aggression = aggression + aggression_result['aggression']
+#     if matches.shape[0] != 0:
+#         aggression_avg = aggression / matches.shape[0]
+#     else:
+#         aggression_avg = 0
+#     return aggression_avg
 
 def calculatePoints(matches, team):
     points = 0
@@ -90,3 +139,18 @@ def calculateGoalDifference(matches, team):
         else:
             goal_diff = goal_diff + (y-x)
     return goal_diff
+
+def calculateColumn(matches, team, column_name):
+    result = 0
+    for index, row in matches.iterrows():
+        if team == row['home_team']:
+            column = row[column_name]
+        else:
+            column = row[column_name]
+        result = result + column
+    if matches.shape[0] != 0:
+        result_avg = result / matches.shape[0]
+    else:
+        result_avg = 0
+    return result_avg
+    

fuzzy.py

@@ -0,0 +1,222 @@
+from simpful import *
+from data_filters import *
+import pandas as pd
+
+FS = FuzzySystem()
+
+# Dominacja OK
+# Jakość strzałów - Witek
+# Agresesywnosc (zolte + czerwone kartki) - Wojtek
+# odbiory i wslizgi (xDef) - Michał, ekspert od xDef
+# statystyki z calego sezonu - Wojtek OK
+# 5 ostatnich spotkan miedzy druzynami - Witek
+def categorizeFuzzyPasses(passes,possession):
+
+
+    FS.set_crisp_output_value("low", 0.0)
+    FS.set_crisp_output_value("average", 0.5)
+    FS.set_crisp_output_value("high", 1.0)
+    
+    Pass1 = TriangleFuzzySet(300,300,600,   term="low") 
+    Pass2 = TriangleFuzzySet(110,440,770,  term="average") 
+    Pass3 = TriangleFuzzySet(200,600,600, term="high")
+
+    FS.add_linguistic_variable("passes", LinguisticVariable([Pass1, Pass2, Pass3], universe_of_discourse=[0,1000]))
+
+    Poss1 = TriangleFuzzySet(30,30,50,   term="low") 
+    Poss2 = TriangleFuzzySet(35,50,65,  term="average") 
+    Poss3 = TriangleFuzzySet(50,70,70, term="high")
+
+    FS.add_linguistic_variable("possession", LinguisticVariable([Poss1, Poss2, Poss3], universe_of_discourse=[0,100]))
+
+    
+    #Pass_domination1 = TriangleFuzzySet(2,2,6,   term="low") 
+    #Pass_domination2 = TriangleFuzzySet(3,5,7,  term="average") 
+    #Pass_domination3 = TriangleFuzzySet(4,8,8, term="high")
+
+    #FS.add_linguistic_variable("passes_domination", LinguisticVariable([Pass_domination1, Pass_domination2, Pass_domination3], universe_of_discourse=[0,10]))
+
+
+    FS.add_rules([ 
+
+    "IF (passes IS low) AND (possession IS low) THEN (pass_domination IS low)", 
+
+    "IF (passes IS high) AND (possession IS high) THEN (pass_domination IS high)",
+
+    "IF (passes IS average) AND (possession IS average) THEN (pass_domination IS average)",
+
+    "IF (passes IS low) AND (possession IS high) THEN (pass_domination IS average)",
+
+    "IF (passes IS high ) AND (possession IS low) THEN (pass_domination IS average)",
+
+    "IF (passes IS average) AND (possession IS high) THEN (pass_domination IS high)",
+
+    "IF (passes IS high) AND (possession IS average) THEN (pass_domination IS high)",
+
+    "IF (passes IS low) AND (possession IS average) THEN (pass_domination IS low)",
+
+    "IF (passes IS average) AND (possession IS low) THEN (pass_domination IS average)"
+    ])  
+
+    FS.set_variable("passes", passes)  
+
+    FS.set_variable("possession", possession)  
+
+    pass_domination = FS.inference() 
+
+    return pass_domination['pass_domination']
+
+
+def categorizeFuzzyDefence(tackles,clearances):
+
+
+    FS.set_crisp_output_value("low", 0.0)
+    FS.set_crisp_output_value("average", 0.5)
+    FS.set_crisp_output_value("high", 1.0)
+    
+    Tackle1 = TriangleFuzzySet(12,12,24,   term="low") 
+    Tackle2 = TriangleFuzzySet(6,18,30,  term="average") 
+    Tackle3 = TriangleFuzzySet(12,24,24, term="high")
+
+    FS.add_linguistic_variable("tackles", LinguisticVariable([Tackle1, Tackle2, Tackle3], universe_of_discourse=[0,80]))
+
+    Clear1 = TriangleFuzzySet(12,12,36,   term="low") 
+    Clear2 = TriangleFuzzySet(9,27,45,  term="average") 
+    Clear3 = TriangleFuzzySet(18,42,42, term="high")
+
+    FS.add_linguistic_variable("clearances", LinguisticVariable([Clear1, Clear2, Clear3], universe_of_discourse=[0,80]))
+
+    FS.add_rules([ 
+
+    "IF (tackles IS low) AND (clearances IS low) THEN (defence_actions IS low)", 
+
+    "IF (tackles IS high) AND (clearances IS high) THEN (defence_actions IS high)",
+
+    "IF (tackles IS average) AND (clearances IS average) THEN (defence_actions IS average)",
+
+    "IF (tackles IS low) AND (clearances IS high) THEN (defence_actions IS average)",
+
+    "IF (tackles IS high ) AND (clearances IS low) THEN (defence_actions IS average)",
+
+    "IF (tackles IS average) AND (clearances IS high) THEN (defence_actions IS high)",
+
+    "IF (tackles IS high) AND (clearances IS average) THEN (defence_actions IS high)",
+
+    "IF (tackles IS low) AND (clearances IS average) THEN (defence_actions IS low)",
+
+    "IF (tackles IS average) AND (clearances IS low) THEN (defence_actions IS average)"
+    ])  
+
+    FS.set_variable("tackles", tackles)  
+
+    FS.set_variable("clearances", clearances)  
+
+    defence_actions = FS.inference() 
+
+    return defence_actions['defence_actions']
+
+def categorizeFuzzyShots(shots_overall, shots_on_target):
+
+    FS.set_crisp_output_value("low", 0.0)
+    FS.set_crisp_output_value("average", 0.5)
+    FS.set_crisp_output_value("high", 1.0)
+
+    Shot_ov1 = TriangleFuzzySet(3,3,12,   term="low") #pozmieniać przedziały (nakładają się)
+    Shot_ov2 = TriangleFuzzySet(2,12,22,  term="average") 
+    Shot_ov3 = TriangleFuzzySet(6,25,25, term="high")
+
+    FS.add_linguistic_variable("shots_overall", LinguisticVariable([Shot_ov1, Shot_ov2, Shot_ov3], universe_of_discourse=[0,35])) 
+
+    Shot_ont1 = TriangleFuzzySet(1,1,6,   term="low") 
+    Shot_ont2 = TriangleFuzzySet(2,5,8,  term="average") 
+    Shot_ont3 = TriangleFuzzySet(3,10,10, term="high")
+
+    FS.add_linguistic_variable("shots_on_target", LinguisticVariable([Shot_ont1, Shot_ont2, Shot_ont3], universe_of_discourse=[0,15])) 
+    
+    
+    #Qual_of_shots1 = TriangleFuzzySet(0,0,0.3,   term="low") 
+    #Qual_of_shots2 = TriangleFuzzySet(0.2,0.5,0.8,  term="medium") 
+    #Qual_of_shots3 = TriangleFuzzySet(0.7,1,1, term="high")
+
+    #FS.add_linguistic_variable("expected_goals", LinguisticVariable([Qual_of_shots1, Qual_of_shots2, Qual_of_shots3], universe_of_discourse=[0,1])) 
+
+    
+    FS.add_rules([ 
+
+    "IF (shots_overall IS low) AND (shots_on_target IS low) THEN (quality_of_shots IS low)", 
+
+    "IF (shots_overall IS high) AND (shots_on_target IS high) THEN (quality_of_shots IS high)",
+
+    "IF (shots_overall IS average) AND (shots_on_target IS average) THEN (quality_of_shots IS average)",
+
+    "IF (shots_overall IS low) AND (shots_on_target IS high) THEN (quality_of_shots IS average)",
+
+    "IF (shots_overall IS high ) AND (shots_on_target IS low) THEN (quality_of_shots IS low)",
+
+    "IF (shots_overall IS average) AND (shots_on_target IS high) THEN (quality_of_shots IS high)",
+
+    "IF (shots_overall IS high) AND (shots_on_target IS average) THEN (quality_of_shots IS average)",
+
+    "IF (shots_overall IS low) AND (shots_on_target IS average) THEN (quality_of_shots IS average)",
+
+    "IF (shots_overall IS average) AND (shots_on_target IS low) THEN (quality_of_shots IS low)"
+
+    ])  
+
+    FS.set_variable("shots_overall", shots_overall)  
+
+    FS.set_variable("shots_on_target", shots_on_target)  
+
+    quality_of_shots = FS.inference() 
+
+    return quality_of_shots['quality_of_shots']
+
+def calculateFuzzyAggression(yellow_cards, red_cards):
+    FS.set_crisp_output_value("low", 0.0)
+    FS.set_crisp_output_value("average", 0.5)
+    FS.set_crisp_output_value("high", 1.0)
+
+    Yellow_cards1 = TriangleFuzzySet(0, 2, 4, term="low")
+    Yellow_cards2 = TriangleFuzzySet(1, 3, 5, term="average")
+    Yellow_cards3 = TriangleFuzzySet(2, 4, 4, term="high")
+
+    FS.add_linguistic_variable("yellow_cards", LinguisticVariable([Yellow_cards1, Yellow_cards2, Yellow_cards3], universe_of_discourse=[0, 10]))
+
+    Red_cards1 = TriangleFuzzySet(0, 0, 1, term="low")
+    Red_cards2 = TriangleFuzzySet(0, 1, 2, term="average")
+    Red_cards3 = TriangleFuzzySet(1, 2, 2, term="high")
+
+    FS.add_linguistic_variable("red_cards", LinguisticVariable([Red_cards1, Red_cards2, Red_cards3], universe_of_discourse=[0, 4]))
+
+    # Pass_domination1 = TriangleFuzzySet(2,2,6,   term="low")
+    # Pass_domination2 = TriangleFuzzySet(3,5,7,  term="average")
+    # Pass_domination3 = TriangleFuzzySet(4,8,8, term="high")
+
+    # FS.add_linguistic_variable("passes_domination", LinguisticVariable([Pass_domination1, Pass_domination2, Pass_domination3], universe_of_discourse=[0,10]))
+
+    FS.add_rules([
+
+        "IF (yellow_cards IS low) AND (red_cards IS low) THEN (aggression IS low)",
+
+        "IF (yellow_cards IS high) AND (red_cards IS high) THEN (aggression IS high)",
+
+        "IF (yellow_cards IS average) AND (red_cards IS average) THEN (aggression IS average)",
+
+        "IF (yellow_cards IS low) AND (red_cards IS high) THEN (aggression IS high)",
+
+        "IF (yellow_cards IS high ) AND (red_cards IS low) THEN (aggression IS high)",
+
+        "IF (yellow_cards IS average) AND (red_cards IS high) THEN (aggression IS high)",
+
+        "IF (yellow_cards IS high) AND (red_cards IS average) THEN (aggression IS high)",
+
+        "IF (yellow_cards IS low) AND (red_cards IS average) THEN (aggression IS average)",
+
+        "IF (yellow_cards IS average) AND (red_cards IS low) THEN (aggression IS average)"
+    ])
+
+    FS.set_variable("yellow_cards", yellow_cards)
+    FS.set_variable("red_cards", red_cards)
+
+    aggression = FS.inference()
+    return aggression['aggression']

main.py

@@ -2,13 +2,13 @@ import pandas as pd
 from simpful import *
 from rules import *
 from data_filters import *
+from fuzzy import *
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.model_selection import train_test_split
 from sklearn.metrics import accuracy_score
 from sklearn.preprocessing import LabelEncoder
 from sklearn.metrics import classification_report
-
+from sklearn.ensemble import GradientBoostingClassifier
-
 
 
 # Ostatnie 5 spotkań
@@ -22,18 +22,25 @@ from sklearn.metrics import classification_report
 
 if __name__ == "__main__":
 
-    df = pd.read_csv('df_full_premierleague.csv')
 
-    result = last5Matches('10/11', 'Stoke City', '2010-10-02', df)
+    df = pd.read_csv('df_parts.csv')
+
+    
+    '''
+
+        
+    df = pd.read_csv('df_full_premierleague.csv')
+    result = last5Matches('10/11', 'Stoke City', '2010-10-02', df)[0]
     #print(result.to_markdown())
     #print(result)
-    result = last5Matches('10/11', 'Blackburn Rovers', '2010-10-02', df)
+    result = last5Matches('10/11', 'Blackburn Rovers', '2010-10-02', df)[0]
     #print(result.to_markdown())
     #print(result)
 
     print(calculatePoints(result,'Blackburn Rovers'))
     print(calculateGoalDifference(result, 'Blackburn Rovers'))
 
+
    # df = generateTrainingData(df)
   #  df = add_column(df, categorize_passes, "c_away_passes", "away_passes")
   #  df = add_column(df, categorize_passes, "c_home_passes", "home_passes")
@@ -44,19 +51,160 @@ if __name__ == "__main__":
    # df = add_column(df, categorize_shots, "c_away_shots", "away_shots")
    # df = add_column(df, categorize_shots, "c_home_shots", "home_shots")
    # print(df.columns)
+   ###############################################################################################
+    # df = add_column(df, getColumnMethod(df, True, 'home_yellow_cards', seasonMatches), "c_home_yellow_cards_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_yellow_cards', seasonMatches), "c_away_yellow_cards_s")
+    # df = add_column(df, getColumnMethod(df, True, 'home_red_cards', seasonMatches), "c_home_red_cards_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_red_cards', seasonMatches), "c_away_red_cards_s")
+    df = df.sort_values(by='date', ascending=False)
+
+    df = add_column(df, getColumnMethod(df, True, 'home_yellow_cards', last5Matches), "c_home_yellow_cards_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_yellow_cards', last5Matches), "c_away_yellow_cards_5m")
+    df = add_column(df, getColumnMethod(df, True, 'home_red_cards', last5Matches), "c_home_red_cards_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_red_cards', last5Matches), "c_away_red_cards_5m")
+
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_yellow_cards', last5MatchesBtwTeams),
+                    "c_home_yellow_cards_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_yellow_cards', last5MatchesBtwTeams),
+                    "c_away_yellow_cards_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_red_cards', last5MatchesBtwTeams), "c_home_red_cards_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_red_cards', last5MatchesBtwTeams), "c_away_red_cards_5btw")
+
+
+
+    ###################################################################################################################
+    # df = add_column(df, getColumnMethod(df, True, 'home_shots', seasonMatches), "c_home_shots_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_shots', seasonMatches), "c_away_shots_s")
+    # df = add_column(df, getColumnMethod(df, True, 'home_shots_on_target', seasonMatches), "c_home_shots_on_target_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_shots_on_target', seasonMatches), "c_away_shots_on_target_s")
+
+    df = add_column(df, getColumnMethod(df, True, 'home_shots', last5Matches), "c_home_shots_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_shots', last5Matches), "c_away_shots_5m")
+    df = add_column(df, getColumnMethod(df, True, 'home_shots_on_target', last5Matches), "c_home_shots_on_target_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_shots_on_target', last5Matches), "c_away_shots_on_target_5m")
+
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_shots', last5MatchesBtwTeams), "c_home_shots")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_shots', last5MatchesBtwTeams), "c_away_shots")
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_shots_on_target', last5MatchesBtwTeams),
+                    "c_home_shots_on_target_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_shots_on_target', last5MatchesBtwTeams),
+                    "c_away_shots_on_target_5btw")
+
+
+    ###################################################################################################################################
+    # df = add_column(df, getColumnMethod(df, True, 'home_tackles', seasonMatches), "c_home_tackles_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_tackles', seasonMatches), "c_away_tackles_s")
+    # df = add_column(df, getColumnMethod(df, True, 'home_clearances', seasonMatches), "c_home_clearances_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_clearances', seasonMatches), "c_away_clearances_s")
+
+    df = add_column(df, getColumnMethod(df, True, 'home_tackles', last5Matches), "c_home_tackles_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_tackles', last5Matches), "c_away_tackles_5m")
+    df = add_column(df, getColumnMethod(df, True, 'home_clearances', last5Matches), "c_home_clearances_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_clearances', last5Matches), "c_away_clearances_5m")
+
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_tackles', last5MatchesBtwTeams), "c_home_tackles_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_tackles', last5MatchesBtwTeams), "c_away_tackles_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_clearances', last5MatchesBtwTeams),
+                    "c_home_clearances_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_clearances', last5MatchesBtwTeams),
+                    "c_away_clearances_5btw")
+
+
+
+    ####################################################################################################################################
+    # df = add_column(df, getColumnMethod(df, True, 'home_passes', seasonMatches), "c_home_passes_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_passes', seasonMatches), "c_away_passes_s")
+    # df = add_column(df, getColumnMethod(df, True, 'home_possession', seasonMatches), "c_home_possession_s")
+    # df = add_column(df, getColumnMethod(df, True, 'away_possession', seasonMatches), "c_away_possession_s")
+
+    df = add_column(df, getColumnMethod(df, True, 'home_passes', last5Matches), "c_home_passes_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_passes', last5Matches), "c_away_passes_5m")
+    df = add_column(df, getColumnMethod(df, True, 'home_possession', last5Matches), "c_home_possession_5m")
+    df = add_column(df, getColumnMethod(df, False, 'away_possession', last5Matches), "c_away_possession_5m")
+
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_passes', last5MatchesBtwTeams), "c_home_passes_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_passes', last5MatchesBtwTeams), "c_away_passes_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, True, 'home_possession', last5MatchesBtwTeams),
+                    "c_home_possession_5btw")
+    df = add_column(df, getColumnMethod5Btw(df, False, 'away_possession', last5MatchesBtwTeams),
+                    "c_away_possession_5btw")
+
+    # TU
+
+
+    ##########################################################################################################################################
 
     df = add_column(df, get_method(df, True, categorize_points, last5Matches), "c_home_form_5m")
     df = add_column(df, get_method(df, False, categorize_points, last5Matches), "c_away_form_5m")
 
-    df = add_column(df, get_method(df, True, categorize_diff, last5Matches), "c_home_diff_5m")#categorize_diff
+    df = add_column(df, get_method(df, True, categorize_diff, seasonMatches), "c_home_diff_5m")
-    df = add_column(df, get_method(df, False, categorize_diff,last5Matches), "c_away_diff_5m")
+    df = add_column(df, get_method(df, False, categorize_diff, seasonMatches), "c_away_diff_5m")
 
-    df = add_column(df, get_method(df, True, categorize_points, seasonMatches), "c_home_form_season")
+    df = add_column(df, get_method(df, True, categorize_points_Btw, last5MatchesBtwTeams), "c_home_form_5btw")
-    df = add_column(df, get_method(df, False, categorize_points, seasonMatches), "c_away_form_season")
+    df = add_column(df, get_method(df, False, categorize_points_Btw, last5MatchesBtwTeams), "c_away_form_5btw")
 
-    df = add_column(df, get_method(df, True, categorize_diff, seasonMatches), "c_home_diff_season")#categorize_diff
+    df = add_column(df, get_method(df, True, categorize_diff_Btw, last5MatchesBtwTeams), "c_home_diff_5btw")
-    df = add_column(df, get_method(df, False, categorize_diff,seasonMatches), "c_away_diff_season")
+    df = add_column(df, get_method(df, False, categorize_diff_Btw, last5MatchesBtwTeams), "c_away_diff_5btw")
 
+    df = add_column(df, get_method(df, True, categorize_points, seasonMatches), "c_home_form_5s")
+    df = add_column(df, get_method(df, False, categorize_points, seasonMatches), "c_away_form_5s")
+
+    df = add_column(df, get_method(df, True, categorize_diff, seasonMatches), "c_home_diff_5s")
+    df = add_column(df, get_method(df, False, categorize_diff, seasonMatches), "c_away_diff_5s")
+
+    df.to_csv('df.csv', index=False)
+#TU sie zapisuje zbior
+    
+    part_size = 50
+    for part in split_to_parts(df, part_size):
+
+
+        part = add_column(part,
+                        getFuzzyMethod(part, calculateFuzzyAggression, True, "c_home_yellow_cards_5m", "c_home_red_cards_5m"),
+                        "c_home_aggression_5m")
+        part = add_column(part, getFuzzyMethod(part, calculateFuzzyAggression, False, "c_away_yellow_cards_5m",
+                                        "c_away_red_cards_5m"), "c_away_aggression_5m")
+
+        part = add_column(part, getFuzzyMethod(part, calculateFuzzyAggression, True, "c_home_yellow_cards_5btw",
+                                        "c_home_red_cards_5btw"), "c_home_aggression_5btw")
+        part = add_column(part, getFuzzyMethod(part, calculateFuzzyAggression, False, "c_away_yellow_cards_5btw",
+                                        "c_away_red_cards_5btw"), "c_away_aggression_5btw")
+
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyShots, True, "c_home_shots_5m", "c_home_shots_5m"),
+                        "c_home_shots_5m")
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyShots, False, "c_away_shots_5m", "c_away_shots_5m"),
+                        "c_away_shots_5m")
+
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyShots, True, "c_home_shots_on_target_5btw",
+                                        "c_home_shots_on_target_5btw"), "c_home_shots_5btw")
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyShots, False, "c_away_shots_on_target_5btw",
+                                        "c_away_shots_on_target_5btw"), "c_away_shots_5btw")
+
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyDefence, True, "c_home_tackles_5m", "c_home_clearances_5m"),
+                        "c_home_defence_5m")
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyDefence, False, "c_away_tackles_5m", "c_away_clearances_5m"),
+                        "c_away_defence_5m")
+
+        part = add_column(part,
+                        getFuzzyMethod(part, categorizeFuzzyDefence, True, "c_home_tackles_5btw", "c_home_clearances_5btw"),
+                        "c_home_defence_5btw")
+        part = add_column(part,
+                        getFuzzyMethod(part, categorizeFuzzyDefence, False, "c_away_tackles_5btw", "c_away_clearances_5btw"),
+                        "c_away_defence_5btw")
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyPasses, True, "c_home_passes_5m", "c_home_possession_5m"),
+                        "c_home_passing_5m")
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyPasses, False, "c_away_passes_5m", "c_away_possession_5m"),
+                        "c_away_passing_5m")
+
+        part = add_column(part, getFuzzyMethod(part, categorizeFuzzyPasses, True, "c_home_passes_5btw", "c_home_possession_5btw"),
+                        "c_home_passing_5btw")
+        part = add_column(part,
+                        getFuzzyMethod(part, categorizeFuzzyPasses, False, "c_away_passes_5btw", "c_away_possession_5btw"),
+                        "c_away_passing_5btw")
+        
+
+        save_to_csv("df_parts", part)
+'''
     df = generateFuzzyLogicData(df)
 
     label_encoder = LabelEncoder()
@@ -68,9 +216,12 @@ if __name__ == "__main__":
     df[['home_team', 'away_team']] = temp.unstack()
     X = df.drop(['result_full', 'date', 'c_home_result', 'c_away_result'], axis=1)
     y = df['c_home_result']
+    #y = label_encoder.fit_transform(df['c_home_result'])
     X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
 
-    model = RandomForestClassifier(n_estimators=100, random_state=42)
+    model = RandomForestClassifier(n_estimators=500, random_state=42)
+    #model = GradientBoostingClassifier(learning_rate=0.1, n_estimators=100, random_state = 42)
+
     model.fit(X_train, y_train)
 
     y_pred = model.predict(X_test)
@@ -81,9 +232,29 @@ if __name__ == "__main__":
     accuracy = accuracy_score(y_test, y_pred)
     print(f'Dokładność modelu: {accuracy}')
     print(classification_report(y_test, y_pred))
-    
-    result = last5Matches('10/11', 'Manchester United', '2010-12-16', df)
-    print(calculatePoints(result,'Manchester United'))
-    print(calculateGoalDifference(result, 'Manchester United'))
 
-    print(categorize_fuzzy_passes(450,50))
+    #print(model.feature_importances_)
+    
+
+    #print(categorize_fuzzy_passes(450,50))
+
+    #df.to_csv('df.csv', index=False)
+
+
+    # df = add_column(df, get_method(df, True, categorize_aggression, last5Matches), "c_home_aggression_5m")#categorize_diff
+    # df = add_column(df, get_method(df, False, categorize_aggression,last5Matches), "c_away_aggression_5m")
+   # df = add_column(df, get_method(df, True, categorize_points, seasonMatches), "c_home_form_season")
+   # df = add_column(df, get_method(df, False, categorize_points, seasonMatches), "c_away_form_season")
+
+    #df = add_column(df, get_method(df, True, categorize_diff, seasonMatches), "c_home_diff_season")#categorize_diff
+   # df = add_column(df, get_method(df, False, categorize_diff,seasonMatches), "c_away_diff_season")
+
+   # df = add_column(df, get_method(df, True, categorize_aggression, seasonMatches), "c_home_aggression_season")#categorize_diff
+   # df = add_column(df, get_method(df, False, categorize_aggression,seasonMatches), "c_away_aggression_season")
+    
+
+    #df = add_column(df, get_method(df, True, categorize_points, last5MatchesBtwTeams), "c_home_form_5btw")
+    #df = add_column(df, get_method(df, False, categorize_points, last5MatchesBtwTeams), "c_away_form_5btw")
+
+    #df = add_column(df, get_method(df, True, categorize_diff, last5MatchesBtwTeams), "c_home_diff_5btw")#categorize_diff
+    #df = add_column(df, get_method(df, False, categorize_diff,last5MatchesBtwTeams), "c_away_diff_5btw")

rules.py

@@ -1,121 +1,6 @@
-import simpful
 from data_filters import *
 import pandas as pd
 
-FS = FuzzySystem()
-
-# Dominacja OK
-# Jakość strzałów - Witek
-# Agresesywnosc (zolte + czerwone kartki) - Wojtek
-# odbiory i wslizgi (xDef) - Michał, ekspert od xDef
-# statystyki z calego sezonu - Wojtek OK
-# 5 ostatnich spotkan miedzy druzynami - Witek
-def categorize_fuzzy_passes(passes,possession):
-
-
-    FS.set_crisp_output_value("low", 0.0)
-    FS.set_crisp_output_value("average", 0.5)
-    FS.set_crisp_output_value("high", 1.0)
-    
-    Pass1 = TriangleFuzzySet(300,300,500,   term="low") 
-    Pass2 = TriangleFuzzySet(300,450,600,  term="average") 
-    Pass3 = TriangleFuzzySet(400,600,600, term="high")
-
-    FS.add_linguistic_variable("passes", LinguisticVariable([Pass1, Pass2, Pass3], universe_of_discourse=[0,1000]))
-
-    Poss1 = TriangleFuzzySet(30,30,45,   term="low") 
-    Poss2 = TriangleFuzzySet(40,50,60,  term="average") 
-    Poss3 = TriangleFuzzySet(55,70,70, term="high")
-
-    FS.add_linguistic_variable("possession", LinguisticVariable([Poss1, Poss2, Poss3], universe_of_discourse=[0,100]))
-
-    
-    #Pass_domination1 = TriangleFuzzySet(2,2,6,   term="low") 
-    #Pass_domination2 = TriangleFuzzySet(3,5,7,  term="average") 
-    #Pass_domination3 = TriangleFuzzySet(4,8,8, term="high")
-
-    #FS.add_linguistic_variable("passes_domination", LinguisticVariable([Pass_domination1, Pass_domination2, Pass_domination3], universe_of_discourse=[0,10]))
-
-
-    FS.add_rules([ 
-
-    "IF (passes IS low) AND (possession IS low) THEN (pass_domination IS low)", 
-
-    "IF (passes IS high) AND (possession IS high) THEN (pass_domination IS high)",
-
-    "IF (passes IS average) AND (possession IS average) THEN (pass_domination IS average)",
-
-    "IF (passes IS low) AND (possession IS high) THEN (pass_domination IS average)",
-
-    "IF (passes IS high ) AND (possession IS low) THEN (pass_domination IS average)",
-
-    "IF (passes IS average) AND (possession IS high) THEN (pass_domination IS high)",
-
-    "IF (passes IS high) AND (possession IS average) THEN (pass_domination IS high)",
-
-    "IF (passes IS low) AND (possession IS average) THEN (pass_domination IS low)",
-
-    "IF (passes IS average) AND (possession IS low) THEN (pass_domination IS average)"
-    ])  
-
-    FS.set_variable("passes", passes)  
-
-    FS.set_variable("possession", possession)  
-
-    pass_domination = FS.inference() 
-
-    return pass_domination
-
-def categorize_fuzzy_shots(shots_overall, shots_on_target):
-
-    FS.set_crisp_output_value("low", 0.0)
-    FS.set_crisp_output_value("average", 0.5)
-    FS.set_crisp_output_value("high", 1.0)
-
-    Shot_ov1 = TriangleFuzzySet(0,0,5,   term="low") #pozmieniać przedziały (nakładają się)
-    Shot_ov2 = TriangleFuzzySet(5,10,15,  term="medium") 
-    Shot_ov3 = TriangleFuzzySet(15,25,25, term="high")
-
-    FS.add_linguistic_variable("shots_overall", LinguisticVariable([Shot_ov1, Shot_ov2, Shot_ov3], universe_of_discourse=[0,35])) 
-
-    Shot_ont1 = TriangleFuzzySet(0,0,2,   term="low") 
-    Shot_ont2 = TriangleFuzzySet(2,4,6,  term="medium") 
-    Shot_ont3 = TriangleFuzzySet(6,10,10, term="high")
-
-    FS.add_linguistic_variable("shots_on_target", LinguisticVariable([Shot_ont1, Shot_ont2, Shot_ont3], universe_of_discourse=[0,15])) 
-    
-    
-    #Qual_of_shots1 = TriangleFuzzySet(0,0,0.3,   term="low") 
-    #Qual_of_shots2 = TriangleFuzzySet(0.2,0.5,0.8,  term="medium") 
-    #Qual_of_shots3 = TriangleFuzzySet(0.7,1,1, term="high")
-
-    #FS.add_linguistic_variable("expected_goals", LinguisticVariable([Qual_of_shots1, Qual_of_shots2, Qual_of_shots3], universe_of_discourse=[0,1])) 
-
-    
-    FS.add_rules([ 
-
-    "IF (shots_overall IS low) AND (shots_on_target IS low) THEN (quality_of_shots IS low)", 
-
-    "IF (shots_overall IS high) AND (shots_on_target IS high) THEN (quality_of_shots IS high)",
-
-    "IF (shots_overall IS average) AND (shots_on_target IS average) THEN (quality_of_shots IS average)",
-
-    "IF (shots_overall IS low) AND (shots_on_target IS high) THEN (quality_of_shots IS high)",
-
-    "IF (shots_overall IS high ) AND (shots_on_target IS low) THEN (quality_of_shots IS low)",
-
-    "IF (shots_overall IS average) AND (possession IS high) THEN (quality_of_shots IS high)",
-
-    "IF (passes IS high) AND (possession IS average) THEN (quality_of_shots IS high)",
-
-    "IF (passes IS low) AND (possession IS average) THEN (quality_of_shots IS low)",
-
-    "IF (passes IS average) AND (possession IS low) THEN (quality_of_shots IS average)"
-
-    ])  
-
-
-
 
 def categorize_shots(shots):
     if shots >= 12:
@@ -143,10 +28,10 @@ def categorize_possesion(shots):
     
 def categorize_points(data, row, teamHome, matches_type):
     if teamHome:
-        data_5 = matches_type(row['season'], row['home_team'], row['date'], data)
+        data_5 = matches_type(row['season'], row['home_team'], row['date'], data)[0]
         points = calculatePoints(data_5,row['home_team'])
     else:
-        data_5 = matches_type(row['season'], row['away_team'], row['date'], data)
+        data_5 = matches_type(row['season'], row['away_team'], row['date'], data)[0]
         points = calculatePoints(data_5,row['away_team'])
     if points <=1:
         return 0
@@ -155,12 +40,30 @@ def categorize_points(data, row, teamHome, matches_type):
     else:
         return 1
 
+def categorize_points_Btw(data, row, teamHome, matches_type):
+    if teamHome:
+        data_5 = matches_type(row['home_team'], row['away_team'], row['date'], data)[0]
+        points = calculatePoints(data_5,row['home_team'])
+    else:
+        data_5 = matches_type(row['home_team'], row['away_team'], row['date'], data)[0]
+        points = calculatePoints(data_5,row['away_team'])
+    if points <=1:
+        return 0
+    elif points >=2:
+        return 2
+    else:
+        return 1
+
+
+
+
 def get_method(data, home_away, method, matches_type):
     values = []
     for index, row in data.iterrows():
         values.append(method(data, row, home_away, matches_type))
     return values
 
+
 def get_points_home(data):
     points = []
     for index, row in data.iterrows():
@@ -180,17 +83,53 @@ def get_points_away(data):
 
 def categorize_diff(data, row, teamHome, matches_type):
     if teamHome:
-        data_5 = matches_type(row['season'], row['home_team'], row['date'], data)
+        data_5 = matches_type(row['season'], row['home_team'], row['date'], data)[0]
         diff = calculateGoalDifference(data_5,row['home_team'])
     else:
-        data_5 = matches_type(row['season'], row['away_team'], row['date'], data)
+        data_5 = matches_type(row['season'], row['away_team'], row['date'], data)[0]
+        diff = calculateGoalDifference(data_5,row['away_team'])
+    if diff <=0:
+        return 0
+    else:
+        return 1
+    
+def categorize_diff_Btw(data, row, teamHome, matches_type):
+    if teamHome:
+        data_5 = matches_type(row['home_team'], row['away_team'], row['date'], data)[0]
+        diff = calculateGoalDifference(data_5,row['home_team'])
+    else:
+        data_5 = matches_type(row['home_team'], row['away_team'], row['date'], data)[0]
+        diff = calculateGoalDifference(data_5,row['away_team'])
+    if diff <=0:
+        return 0
+    else:
+        return 1
+    
+
+def categorize_diff(data, row, teamHome, matches_type):
+    if teamHome:
+        data_5 = matches_type(row['season'], row['home_team'], row['date'], data)[0]
+        diff = calculateGoalDifference(data_5,row['home_team'])
+    else:
+        data_5 = matches_type(row['season'], row['away_team'], row['date'], data)[0]
         diff = calculateGoalDifference(data_5,row['away_team'])
     if diff <=0:
         return 0
     else:
         return 1
 
-
+# def categorize_aggression(data, row, teamHome, matches_type):
+#     if teamHome:
+#         data_5 = matches_type(row['season'], row['home_team'], row['date'], data)
+#         diff = calculateAggression(data_5,row['home_team'])
+#     else:
+#         data_5 = matches_type(row['season'], row['away_team'], row['date'], data)
+#         diff = calculateAggression(data_5,row['away_team'])
+#     return diff
+#    # if diff <=0:
+#     #    return 0
+#    # else:
+#    #     return 1
 def get_diff_home(data):
     points = []
     for index, row in data.iterrows():
@@ -223,3 +162,46 @@ def get_result_list(df, home_team):
     for score in df['result_full']:
         results.append(getResult(score,home_team))
     return results
+
+def getColumnMethod(data, home_away, column, matches_type):
+    values = []
+    for index, row in data.iterrows():
+        values.append(getColumnByMatches(data, row, home_away, matches_type,column))
+    return values
+
+def getColumnMethod5Btw(data, home_away, column, matches_type):
+    values = []
+    for index, row in data.iterrows():
+        values.append(getColumnByMatches5Btw(data, row, home_away, matches_type,column))
+    return values
+
+def getFuzzyMethod(data, method, home_away, frist_column, second_column):
+    values = []
+    for index, row in data.iterrows():
+        values.append(getFuzzyByMatches(method, row, home_away, frist_column, second_column))
+    return values
+
+def getColumnByMatches(data, row, teamHome, matches_type, column):
+    if teamHome:
+        data_5, end = matches_type(row['season'], row['home_team'], row['date'], data)
+        result = calculateColumn(data_5, row['home_team'], column)
+    else:
+        data_5, end = matches_type(row['season'], row['away_team'], row['date'], data)
+        result = calculateColumn(data_5, row['away_team'], column)
+    return result
+
+def getColumnByMatches5Btw(data, row, teamHome, matches_type, column):
+    if teamHome:
+        data_5, end = last5MatchesBtwTeams(row['home_team'], row['away_team'], row['date'], data)
+        result = calculateColumn(data_5, row['home_team'], column)
+    else:
+        data_5, end = matches_type(row['home_team'], row['away_team'], row['date'], data)
+        result = calculateColumn(data_5, row['away_team'], column)
+    return result
+
+def getFuzzyByMatches(method, row, teamHome, frist_column, second_column):
+    if teamHome:
+        result = method(row[frist_column], row[second_column])
+    else:
+        result = method(row[frist_column], row[second_column])
+    return result