done

2026-04-21 19:33:24 +08:00
parent 9a72566753
commit 3268080401
26 changed files with 1816 additions and 318 deletions
@@ -0,0 +1,84 @@
+"""
+批量回跑语义分析 — 验证新prompt效果
+用法: python scripts/reanalyze_sample.py
+"""
+import asyncio
+import sys
+import os
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+from app.services.analysis import analyze_post
+from app.database import AsyncSessionLocal
+from app.models import Post
+from sqlalchemy import select
+
+
+async def main():
+    async with AsyncSessionLocal() as db:
+        result = await db.execute(
+            select(Post)
+            .order_by(Post.published_at.desc())
+            .limit(60)
+        )
+        posts = result.scalars().all()
+
+    print(f"分析 {len(posts)} 条帖子 (model: claude-sonnet-4-6)\n")
+
+    signals = {"buy": 0, "sell": 0, "short": 0, "hold": 0}
+    relevant_count = 0
+
+    correct_m5, checked_m5 = 0, 0
+    correct_m15, checked_m15 = 0, 0
+    correct_m1h, checked_m1h = 0, 0
+
+    for i, post in enumerate(posts):
+        analysis = await analyze_post(post.text)
+        signals[analysis["signal"]] += 1
+        if analysis["relevant"]:
+            relevant_count += 1
+
+        if analysis["relevant"]:
+            sig = analysis["signal"].upper()
+            conf = analysis["confidence"]
+
+            actual_m5  = post.price_impact_m5
+            actual_m15 = post.price_impact_m15
+            actual_m1h = post.price_impact_m1h
+
+            def dir_ok(actual, signal):
+                if actual is None: return None
+                if signal == "buy":   return actual > 0
+                if signal == "short": return actual < 0
+                return None
+
+            m5_ok  = dir_ok(actual_m5,  analysis["signal"])
+            m15_ok = dir_ok(actual_m15, analysis["signal"])
+            m1h_ok = dir_ok(actual_m1h, analysis["signal"])
+
+            if m5_ok  is not None: checked_m5  += 1; correct_m5  += int(m5_ok)
+            if m15_ok is not None: checked_m15 += 1; correct_m15 += int(m15_ok)
+            if m1h_ok is not None: checked_m1h += 1; correct_m1h += int(m1h_ok)
+
+            # accuracy marks
+            marks = ""
+            if m5_ok  is not None: marks += f" 5m:{'✅' if m5_ok  else '❌'}"
+            if m15_ok is not None: marks += f" 15m:{'✅' if m15_ok else '❌'}"
+            if m1h_ok is not None: marks += f" 1h:{'✅' if m1h_ok else '❌'}"
+
+            print(f"[{i+1:03d}] {post.published_at.strftime('%m-%d %H:%M')} | {sig:5s} {conf:3d}% |{marks}")
+            print(f"      {post.text[:110]}")
+            print(f"      → {analysis['reasoning'][:200]}")
+            print()
+
+        await asyncio.sleep(0.2)
+
+    print("=" * 70)
+    print(f"总计: {len(posts)} 帖 | Relevant: {relevant_count} ({relevant_count/len(posts)*100:.0f}%)")
+    print(f"信号: BUY={signals['buy']} SHORT={signals['short']} SELL={signals['sell']} HOLD={signals['hold']}")
+    print()
+    if checked_m5:  print(f"准确率  5m: {correct_m5}/{checked_m5}  = {correct_m5/checked_m5*100:.0f}%")
+    if checked_m15: print(f"准确率 15m: {correct_m15}/{checked_m15} = {correct_m15/checked_m15*100:.0f}%")
+    if checked_m1h: print(f"准确率  1h: {correct_m1h}/{checked_m1h} = {correct_m1h/checked_m1h*100:.0f}%")
+
+if __name__ == "__main__":
+    asyncio.run(main())