first (part 2: other directories)

2026-04-16 15:11:25 +03:00
parent 7814e0bc6b
commit 2f0b85a0c7
203 changed files with 59989 additions and 0 deletions
@@ -0,0 +1,60 @@
+
+import os
+import psycopg2
+from psycopg2.extras import RealDictCursor
+
+def inspect_recoverable_data():
+    try:
+        db_url = os.environ.get('DATABASE_URL', 'postgresql://suggestbet:SuGGesT2026SecuRe@localhost:15432/boilerplate_db')
+        conn = psycopg2.connect(db_url)
+        cursor = conn.cursor(cursor_factory=RealDictCursor)
+        
+        print("\n🔍 RECOVERABLE DATA INSPECTION (Ignoring 'state' column)\n")
+
+        # 1. Total Scored Football Matches
+        # Logic: Valid score and valid date is what matters for training
+        cursor.execute("""
+            SELECT COUNT(*) as count 
+            FROM matches 
+            WHERE sport='football' 
+              AND score_home IS NOT NULL 
+              AND score_away IS NOT NULL
+        """)
+        total_scored = cursor.fetchone()['count']
+        print(f"Total Scored Football Matches: {total_scored:,}")
+
+        if total_scored == 0:
+            return
+
+        # 2. Recoverable Matches with Stats
+        cursor.execute("""
+            SELECT COUNT(DISTINCT m.id) as count 
+            FROM matches m
+            JOIN match_team_stats s ON m.id = s.match_id
+            WHERE m.sport='football' 
+              AND m.score_home IS NOT NULL 
+              AND m.score_away IS NOT NULL
+        """)
+        stats_count = cursor.fetchone()['count']
+        print(f"Matches with Team Stats:       {stats_count:,} ({stats_count/total_scored*100:.1f}%)")
+
+        # 3. Recoverable 'Golden Dataset' (Stats + Odds + Events)
+        cursor.execute("""
+            SELECT COUNT(DISTINCT m.id) as count 
+            FROM matches m
+            JOIN match_team_stats s ON m.id = s.match_id
+            JOIN match_player_events e ON m.id = e.match_id
+            JOIN odd_categories oc ON m.id = oc.match_id
+            WHERE m.sport='football' 
+              AND m.score_home IS NOT NULL 
+              AND m.score_away IS NOT NULL
+        """)
+        golden_count = cursor.fetchone()['count']
+        print(f"\n✨ TRUE GOLDEN DATASET (Scored + All 3): {golden_count:,} ({golden_count/total_scored*100:.1f}%)")
+
+        conn.close()
+    except Exception as e:
+        print(f"Error: {e}")
+
+if __name__ == "__main__":
+    inspect_recoverable_data()