{"id":5119,"date":"2026-10-28T07:20:00","date_gmt":"2026-10-27T23:20:00","guid":{"rendered":"https:\/\/laoyulaoyu.com\/?p=5119"},"modified":"2026-10-23T19:20:41","modified_gmt":"2026-10-23T11:20:41","slug":"1-87%e4%b8%87%e6%98%9f%e7%9a%84%e5%bc%80%e6%ba%90%e9%87%8f%e5%8c%96%e5%b9%b3%e5%8f%b0qbot%ef%bc%8c%e5%9b%9e%e6%b5%8b%e8%b7%91%e5%87%ba%e6%9d%a5%e5%8d%b4%e6%98%af%e4%ba%8f%e7%9a%84","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2026\/10\/28\/1-87%e4%b8%87%e6%98%9f%e7%9a%84%e5%bc%80%e6%ba%90%e9%87%8f%e5%8c%96%e5%b9%b3%e5%8f%b0qbot%ef%bc%8c%e5%9b%9e%e6%b5%8b%e8%b7%91%e5%87%ba%e6%9d%a5%e5%8d%b4%e6%98%af%e4%ba%8f%e7%9a%84\/","title":{"rendered":"1.87\u4e07\u661f\u7684\u5f00\u6e90\u91cf\u5316\u5e73\u53f0Qbot\uff0c\u56de\u6d4b\u8dd1\u51fa\u6765\u5374\u662f\u4e8f\u7684"},"content":{"rendered":"\n

\u4f5c \u8005\uff1a\u8001\u4f59\u635e\u9c7c<\/a><\/p>\n\n\n\n

\u539f\u521b\u4e0d\u6613\uff0c\u8f6c\u8f7d\u8bf7\u6807\u660e\u51fa\u5904\u53ca\u539f\u4f5c\u8005\u3002<\/mark><\/strong><\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\n

\u6280\u672f\u8ba8\u8bba\u4e0d\u6784\u6210\u4efb\u4f55\u64cd\u4f5c\u5efa\u8bae\u3002
\u6a21\u578b\u8bc4\u6d4b\u4e0d\u7b49\u4e8e\u6295\u8d44\u5efa\u8bae\u3002
\u4e0d\u8350\u80a1\u3001\u4e0d\u5356\u8bfe\u3001\u4e0d\u627f\u8bfa\u6536\u76ca\u3002
\u5e02\u573a\u6709\u98ce\u9669\uff0c\u8bf7\u5408\u6cd5\u5408\u89c4\u6295\u8d44\u3002<\/p>\n<\/blockquote>\n\n\n\n

\u672c\u6587\u7ea6 3100 \u5b57 | \u9884\u8ba1\u9605\u8bfb 10 \u5206\u949f\uff0c\u540e\u6709\u4ee3\u7801\u548c\u9879\u76ee\u5730\u5740\uff0c\u5efa\u8bae\u5148\u6536\u85cf<\/p>\n\n\n\n

\n\n\n\n

Qbot \u662f UFund-Me \u5f00\u6e90\u7684\u91cf\u5316\u5e73\u53f0\uff0c\u5728 GitHub \u4e0a\u6709 18718 star\u30012613 fork\u3001MIT \u534f\u8bae\uff0c2022 \u5e74 11 \u6708\u5efa\u7684\u4ed3\u3002<\/p>\n\n\n\n

\u7ecf\u5178\u7b56\u7565\u90a3\u4e00\u652f\u662f\u771f\u80fd\u8dd1\u7684\uff0cAI \u7b56\u7565\u90a3\u4e00\u652f\u662f\u574f\u7684\uff0c\u9700\u8981\u6539\u9020\u3002<\/p>\n\n\n\n

\u6211\u82b1\u4e86\u4e9b\u65f6\u95f4\u628a\u5b83\u62c9\u4e0b\u6765\u3001\u88c5\u8d77\u6765\u3001\u628a\u81ea\u5e26\u811a\u672c\u8dd1\u4e86\u4e00\u904d\uff0c\u53c8\u6252\u4e86\u4e00\u904d\u6e90\u7801\u3002\u6316\u51fa\u6765\u7684\u4e1c\u897f\u6bd4\u9884\u671f\u591a\uff0c\u4e5f\u6709\u51e0\u4e2a\u5730\u65b9\u8ba9\u6211\u76ef\u7740\u5c4f\u5e55\u6123\u4e86\u4e00\u4f1a\u513f\u3002<\/p>\n\n\n\n

\u5b83\u7684\u91ce\u5fc3\uff0c\u4e00\u6761\u94fe\u8def\u8dd1\u5b8c<\/strong><\/h2>\n\n\n\n

Qbot \u60f3\u505a\u7684\u4e8b\u5f88\u76f4\u767d\u3002\u6570\u636e\u83b7\u53d6\u3001\u7b56\u7565\u5f00\u53d1\u3001\u7b56\u7565\u56de\u6d4b\u3001\u6a21\u62df\u4ea4\u6613\u3001\u5b9e\u76d8\u4ea4\u6613\uff0c\u653e\u540c\u4e00\u5957\u6d41\u7a0b\u91cc\u3002<\/p>\n\n\n\n

README \u91cc\u753b\u4e86\u4e00\u5f20\u7ed3\u6784\u56fe\uff0c\u4ece\u4e0a\u5230\u4e0b\u56db\u5c42\u3002\u5e95\u5c42\u662f qlib \u548c\u6df1\u5ea6\u5b66\u4e60\u7b56\u7565\u3002\u4e2d\u95f4\u662f backtrader \u548c easyquant \u4e24\u4e2a\u56de\u6d4b\u5f15\u64ce\uff0c\u4e0a\u5c42\u662f vnpy\u3001pytrader\u3001pyfunds \u4e09\u4e2a\u4ea4\u6613\u6846\u67b6\u3002\u6700\u5916\u5c42\u5957 quantstats \u505a\u7ee9\u6548\u770b\u677f\u3002<\/p>\n\n\n\n

\"\"\/<\/figure>\n\n\n\n

\u4f5c\u8005\u81ea\u5df1\u5199\u4e86\u4e00\u53e5\u5b9a\u4f4d\uff0cAI \u667a\u80fd\u91cf\u5316\u6295\u7814\u5e73\u53f0\u3002\u8981\u8986\u76d6\u80a1\u7968\u3001\u57fa\u91d1\u3001\u671f\u8d27\u3001\u52a0\u5bc6\u8d27\u5e01\u56db\u7c7b\u6807\u7684\u3002<\/p>\n\n\n\n

\u4ed3\u5e93\u91cc\u7684\u7ecf\u5178\u7b56\u7565\u76ee\u5f55\u633a\u5168\u3002\u5747\u7ebf\u3001\u5e03\u6797\u3001\u6d77\u9f9f\u3001\u5c0f\u5e02\u503c\u3001\u591a\u56e0\u5b50\u3001Alpha \u5bf9\u51b2\u3001\u6307\u6570\u589e\u5f3a\u3001RSRS \u62e9\u65f6\uff0c\u914d\u5bf9\u4ea4\u6613\u548c\u9009\u80a1\u90fd\u6709\u5355\u72ec\u7684\u6587\u6863\u3002<\/p>\n\n\n\n

\u76ee\u5f55\u770b\u7740\u662f\u9f50\u7684\u3002<\/p>\n\n\n\n

\u5b9e\u6d4b\uff0c\u5b98\u65b9\u811a\u672c\u8dd1\u901a\u4e86\uff0c\u7ed3\u679c\u662f\u4e8f\u7684<\/strong><\/h2>\n\n\n\n

\u6211\u6311\u4e86\u4ed3\u5e93\u91cc\u6700\u5e72\u51c0\u7684\u4e00\u4e2a\u7b56\u7565\uff0cqbot\/strategies\/boll_strategy_bt.py<\/code>\u300213 \u65e5\u5e03\u6797\u5e26\uff0c\u8dcc\u7834\u4e0b\u8f68\u4e70\uff0c\u7a81\u7834\u4e0a\u8f68\u5356\u3002<\/p>\n\n\n\n

\u6570\u636e\u7528\u62db\u5546\u94f6\u884c 600036\uff0c2021 \u5e74 1 \u6708 4 \u65e5\u5230 2022 \u5e74 12 \u6708 6 \u65e5\uff0c467 \u6839 K \u7ebf\uff0c\u524d\u590d\u6743\u3002\u521d\u59cb\u8d44\u91d1 100 \u4e07\uff0c\u624b\u7eed\u8d39 0.1%\u3002<\/p>\n\n\n\n

\u7b56\u7565\u7c7b\u6211\u662f\u76f4\u63a5\u4ece\u4ed3\u5e93 import \u8fdb\u6765\u7684\uff0c\u4e00\u884c\u90fd\u6ca1\u6539\u3002\u552f\u4e00\u52a8\u7684\u662f\u53d6\u6570\u3002\u811a\u672c\u91cc\u5199\u7684 akshare \u4e1c\u8d22\u63a5\u53e3 stock_zh_a_hist<\/code> \u5728\u6211\u8fd9\u53f0\u673a\u5668\u4e0a\u8fde\u4e0d\u4e0a\uff0c\u6211\u6362\u6210\u65b0\u6d6a\u7684 stock_zh_a_daily<\/code>\u3002<\/p>\n\n\n\n

\u8dd1\u901a\u4e86\u3002\u671f\u672b 999586.53 \u5143\uff0c\u4e24\u5e74\u4e8f\u4e86 413 \u5757\u3002<\/p>\n\n\n\n

\"\"
\u56db\u79cd\u8dd1\u6cd5\u90fd\u8dd1\u540c\u4e00\u6bb5\u884c\u60c5\u3002\u5b98\u65b9\u9ed8\u8ba4\u53c2\u6570\u4e0b\u7684\u56de\u64a4\u53ea\u6709 0.13%\uff0c\u662f\u56e0\u4e3a\u4ed3\u4f4d\u592a\u5c0f\uff0c\u4e0d\u662f\u7b56\u7565\u7a33<\/figcaption><\/figure>\n\n\n\n

\u540c\u671f\u4e00\u76f4\u62ff\u7740\u662f 966275.20 \u5143\uff0c\u4e8f 33725 \u5143\u3002\u80a1\u4ef7\u4ece 32.79 \u8dcc\u5230 29.14\uff0c\u8dcc\u4e86 11.13%\u3002<\/p>\n\n\n\n

\u8fd9\u4e48\u770b\u5e03\u6797\u7b56\u7565\u662f\u8dd1\u8d62\u57fa\u51c6\u7684\u3002\u5b98\u65b9\u9ed8\u8ba4\u53c2\u6570\u4e0b\u6210\u4ea4 6 \u7b14\uff0c3 \u80dc 3 \u8d1f\uff0c\u6700\u5927\u56de\u64a4 0.13%\u3002<\/p>\n\n\n\n

\u4f46\u8fd9\u4e2a\u8868\u683c\u6709\u4e2a\u5730\u65b9\u4e0d\u5bf9\u52b2\uff0c\u6211\u7b2c\u4e00\u773c\u5c31\u6ce8\u610f\u5230\u4e86\uff0c\u8bf4\u4e0d\u597d\u662f\u54ea\u513f\u4e0d\u5bf9\uff0c\u7ffb\u56de\u53bb\u624d\u770b\u660e\u767d\u3002<\/p>\n\n\n\n

\u5b98\u65b9\u811a\u672c\u9ed8\u8ba4\u53ea\u7528\u4e86 0.36% \u7684\u672c\u91d1<\/strong><\/h2>\n\n\n\n

boll_strategy_bt.py<\/code> \u91cc\u5199\u7740 cerebro.addsizer(bt.sizers.FixedSize, stake=100)<\/code>\u3002\u521d\u59cb\u8d44\u91d1 100 \u4e07\u3002<\/p>\n\n\n\n

stake=100 \u7684\u610f\u601d\u662f\u6bcf\u6b21\u56fa\u5b9a\u4e70 100 \u80a1\u3002600036 \u8fd9\u6bb5\u65f6\u95f4\u5747\u4ef7 35.67 \u5143\uff0c100 \u80a1\u5c31\u662f 3567 \u5143\u3002<\/p>\n\n\n\n

100 \u4e07\u672c\u91d1\uff0c3567 \u5143\u5728\u573a\u4e0a\u3002\u5360 0.36%\u3002<\/p>\n\n\n\n

\"\"
100 \u4e07\u672c\u91d1\uff0c\u5b98\u65b9\u811a\u672c\u53ea\u7528\u4e86 3567 \u5143\u5728\u573a\u3002\u653e\u5927\u4ed3\u4f4d\u540e\uff0c\u771f\u5b9e\u76c8\u4e8f\u624d\u663e\u51fa\u6765<\/figcaption><\/figure>\n\n\n\n

\u5269\u4e0b 99.64% \u7684\u94b1\u8eba\u5728\u8d26\u4e0a\uff0c\u4e00\u5206\u94b1\u6ca1\u53c2\u4e0e\u4ea4\u6613\u3002<\/p>\n\n\n\n

\u6240\u4ee5\u5b98\u65b9\u811a\u672c\u8dd1\u51fa\u6765\u90a3\u4e2a\u51e0\u4e4e\u4e0d\u4e8f\u7684\u7ed3\u679c\uff0c\u4e0d\u662f\u7b56\u7565\u6709\u591a\u7a33\uff0c\u662f\u4ed3\u4f4d\u5c0f\u5230\u6536\u76ca\u88ab\u7a00\u91ca\u6210\u4e86\u770b\u4e0d\u89c1\u3002\u540c\u4e00\u4e2a\u7b56\u7565\u6211\u628a stake \u6539\u6210 30000 \u6ee1\u4ed3\uff0c\u51c0\u4e8f\u7acb\u523b\u53d8\u6210 -1.03%\u3002<\/p>\n\n\n\n

\u540c\u4e00\u6b21\u56de\u6d4b\uff0c\u90a3\u4e2a\u6570\u5b57\u770b\u8d77\u6765\u7a33\u5f97\u4e0d\u884c\u3002\u53ef\u8fd9\u662f\u771f\u7684\u5417\uff1f\u6211\u7b2c\u4e00\u6b21\u770b\u5230\u8fd9\u884c\u8f93\u51fa\u7684\u65f6\u5019\uff0c\u5fc3\u91cc\u662f\u6709\u70b9\u4fe1\u7684\u3002<\/p>\n\n\n\n

\u653e\u5927 25 \u500d\u4ed3\u4f4d\uff0c\u4e8f\u635f\u4e5f\u653e\u5927 25 \u500d\u3002\u8fd9\u4e2a\u6570\u6211\u76ef\u4e86\u4e00\u4f1a\u513f\u624d\u6562\u5199\u4e0b\u6765\u3002<\/p>\n\n\n\n

\u8fd9\u4e00\u70b9\u503c\u5f97\u5355\u72ec\u62ce\u51fa\u6765\u8bf4\u3002\u65b0\u624b\u770b\u56de\u6d4b\u8f93\u51fa\u6700\u5bb9\u6613\u6389\u8fd9\u4e2a\u5751\uff0c\u770b\u5230\u4e00\u4e2a\u51e0\u4e4e\u4e0d\u4e8f\u7684\u7ed3\u679c\uff0c\u7b2c\u4e00\u53cd\u5e94\u662f\u7b56\u7565\u7a33\uff0c\u800c\u4e0d\u662f”\u4ed3\u4f4d\u662f\u4e0d\u662f\u592a\u5c0f\u4e86”\u3002<\/p>\n\n\n\n

0.1% \u624b\u7eed\u8d39\uff0c\u4e24\u5e74\u5403\u6389 5154 \u5143<\/strong><\/h2>\n\n\n\n

\u65e2\u7136\u8bf4\u5230\u6210\u672c\uff0c\u6211\u628a\u624b\u7eed\u8d39\u53c2\u6570\u8c03\u6210 0 \u518d\u8dd1\u4e00\u904d\uff0c\u5176\u4ed6\u4e00\u5f8b\u4e0d\u52a8\u3002<\/p>\n\n\n\n

\u96f6\u6210\u672c\u90a3\u7248\u671f\u672b 994900.00\uff0c\u51c0\u4e8f 5100 \u5143\u30020.1% \u624b\u7eed\u8d39\u90a3\u7248\u671f\u672b 989745.70\uff0c\u51c0\u4e8f 10254.30 \u5143\u3002<\/p>\n\n\n\n

\"\"
\u540c\u4e00\u5957\u7b56\u7565\u4ee3\u7801\uff0c\u53ea\u6539\u624b\u7eed\u8d39\u53c2\u6570\u30020.1% \u7684\u5355\u8fb9\u8d39\u7387\uff0c\u4e24\u5e74\u591a\u82b1 5154 \u5143<\/figcaption><\/figure>\n\n\n\n

\u5dee 5154 \u5143\u3002<\/p>\n\n\n\n

\u6ee1\u4ed3\u6761\u4ef6\u4e0b\u7b56\u7565\u4e00\u5171\u6210\u4ea4 3 \u6b21\uff0c1 \u80dc 2 \u8d1f\u3002\u4e5f\u5c31\u662f\u8bf4\u8fd9 10254 \u5143\u7684\u51c0\u4e8f\u91cc\uff0c\u624b\u7eed\u8d39\u5360\u4e86 5154 \u5143\uff0c\u4e00\u534a\u3002<\/p>\n\n\n\n

3 \u6b21\u6210\u4ea4\u5403 5154 \u5143\u624b\u7eed\u8d39\uff0c\u662f\u56e0\u4e3a\u6ee1\u4ed3\u4e00\u7b14\u5c31\u662f 30 \u4e07\u672c\u91d1\u30020.1% \u4e00\u6b21\u8fdb\u51fa 600 \u5143\uff0c3 \u6b21\u6765\u56de 3600\uff0c\u5269\u4e0b\u7684\u662f\u5f00\u4ed3\u90a3\u4e00\u6b21\u3002<\/p>\n\n\n\n

\u6240\u4ee5 Qbot \u8fd9\u4e2a\u7b56\u7565\u5728\u6ee1\u4ed3\u65f6\u51c0\u4e8f\uff0c\u6210\u672c\u548c\u7b56\u7565\u4e8f\u635f\u5dee\u4e0d\u591a\u5404\u5360\u4e00\u534a\u3002\u8fd9\u4e0d\u7b97\u81f4\u547d\u3002\u95ee\u9898\u662f\u5b98\u65b9\u811a\u672c\u7684\u9ed8\u8ba4\u8f93\u51fa\u6839\u672c\u770b\u4e0d\u51fa\u6210\u672c\u7684\u5b58\u5728\uff0c\u4ed3\u4f4d\u592a\u5c0f\uff0c\u624b\u7eed\u8d39\u88ab\u7a00\u91ca\u5230\u770b\u4e0d\u89c1\u3002<\/p>\n\n\n\n

\u4e09\u5904\u786c\u4f24\uff0c\u90fd\u5728 AI \u7b56\u7565\u90a3\u4e00\u652f<\/strong><\/h2>\n\n\n\n

\u88c5\u4f9d\u8d56\u7684\u65f6\u5019\u6211\u5c31\u89c9\u5f97\u4e0d\u5bf9\u3002\u5b98\u65b9 README \u8ba9\u4f60\u8dd1 pip install -r dev\/requirements.txt<\/code>\u3002<\/p>\n\n\n\n

\u6211\u5148\u5728 Python 3.13 \u4e0a\u8bd5\uff0c\u79d2\u5d29\u3002<\/p>\n\n\n\n

AttributeError: module 'configparser' has no attribute 'SafeConfigParser'<\/code><\/pre>\n\n\n\n
\u8fd9\u662f pyfolio<\/code> \u90a3\u4e2a\u5305\u7684\u95ee\u9898\u3002quantopian \u65e9\u5173\u505c\u4e86\uff0c\u4ed3\u5e93\u8fd8\u6302\u7740\uff0c\u5b83\u7684 setup.py \u91cc\u7528\u4e86 Python 3.12 \u5220\u6389\u7684 API\u3002<\/p>\n\n\n\n
\u6362\u6210 Python 3.9\uff0cREADME \u58f0\u660e\u652f\u6301\u7684\u7248\u672c\uff0cpyfolio \u8fd9\u4e2a\u9519\u8fc7\u4e86\u3002\u4f46\u4e0b\u4e00\u4e2a\u5751\u63a5\u4e0a\u6765\u3002matplotlib==3.2.2<\/code> \u8fd9\u4e2a\u9489\u6b7b\u7684\u7248\u672c\uff0c\u5728 Apple Silicon \u4e0a\u6ca1\u6709\u9884\u7f16\u8bd1\u5305\u3002\u5f97\u4e0b 40MB \u6e90\u7801\u81ea\u5df1\u7f16\u3002<\/p>\n\n\n\n
\u6211\u6700\u540e\u662f\u628a\u8fd9\u51e0\u4e2a\u574f\u5305\u5254\u6389\uff0c\u7528\u5269\u4e0b\u7684\u4f9d\u8d56\u628a\u73af\u5883\u8dd1\u8d77\u6765\u7684\u3002<\/p>\n\n\n\n
\u7136\u540e\u6211\u6838\u5bf9\u4e86\u4e00\u904d requirements \u548c\u4ee3\u7801\u91cc\u5b9e\u9645 import \u7684\u4e1c\u897f\u3002<\/p>\n\n\n\n
tushare\u3001keras\u3001tensorflow\u3001rlkit\u3001easyquant\u3001qlib\u3001vnpy\u3001lightgbm\u3001gym\u3001ta-lib\u3001wxpython\u3002\u5341\u4e00\u4e2a\u5305\uff0c\u4e00\u4e2a\u90fd\u4e0d\u5728 requirements.txt \u91cc\u3002<\/p>\n\n\n\n
\u8fd9\u5341\u4e00\u4e2a\u5305\uff0c\u5168\u662f\u4ed3\u5e93\u81ea\u5e26\u7b56\u7565\u4ee3\u7801 import<\/code> \u7684\u3002<\/p>\n\n\n\n
\u8fd9\u89e3\u91ca\u4e86\u4e00\u4e2a\u73b0\u8c61\u3002\u4e3a\u4ec0\u4e48\u8fd9\u4e2a\u4ed3\u5e93 star \u6da8\u5230 1.8 \u4e07\uff0c\u6211\u4e00\u5f00\u59cb\u60f3\u4e0d\u660e\u767d\uff0c\u4f46 issue \u91cc”\u88c5\u4e0d\u4e0a””\u62a5 xxx \u9519”\u80fd\u6392\u5230\u524d\u5341\u300276 \u4e2a issue \u91cc\u6709 10 \u4e2a\u662f\u5b89\u88c5\u76f8\u5173\u3002\u8bc4\u8bba\u6700\u591a\u7684 #10<\/a>\uff0c\u6807\u9898\u662f\u300cmac python3.9 pip install -r requirements.txt error\u300d\uff0c\u5e95\u4e0b\u6709 11 \u6761\u8ba8\u8bba\u3002<\/p>\n\n\n\n
\u88c5\u5f97\u4e0a\u7684\u4eba\u662f\u628a\u7f3a\u7684\u4e1c\u897f\u81ea\u5df1\u8865\u9f50\u4e86\u3002\u88c5\u4e0d\u4e0a\u7684\u4eba\u5c31\u4e00\u76f4\u5361\u5728\u90a3\u3002<\/p>\n\n\n\n
\u5f3a\u5316\u5b66\u4e60\u7b56\u7565\u5f15\u7528\u4e86\u4e00\u4e2a\u4e0d\u5b58\u5728\u7684\u4e1c\u897f<\/strong><\/h2>\n\n\n\n
qbot\/strategies\/rl_strategy_bt.py<\/code> \u8fd9\u4e2a\u6587\u4ef6\uff0c\u6211\u7528 AST \u9759\u6001\u626b\u4e86\u4e00\u904d\u6240\u6709 self \u5c5e\u6027\u5f15\u7528\u3002<\/p>\n\n\n\n
\u7b2c 48 \u884c\u5199\u7684\u662f action = self.model.predict(self.obs.reshape(1, -1))[0]<\/code><\/p>\n\n\n\n
self.model \u4ece\u54ea\u6765\uff1f\u6211\u628a\u6574\u4e2a\u7c7b\u626b\u4e86\u4e00\u904d\uff0c\u6ca1\u6709\u3002\u8fd9\u4e2a\u7c7b\u4ece\u6765\u6ca1\u521b\u5efa\u8fc7\u6a21\u578b\u3002<\/p>\n\n\n\n
\u7b2c 57 \u884c\u5199\u7684\u662f self.log('(MA Period %2d) Ending Value %.2f' % ...)<\/code><\/p>\n\n\n\n
self.log \u4e5f\u6ca1\u5b9a\u4e49\u3002\u8fd9\u4e2a\u6587\u4ef6\u91cc\u6ca1\u6709 log \u65b9\u6cd5\u3002<\/p>\n\n\n\n
\u7b2c 58 \u884c\u5f15\u7528\u4e86 self.params.maperiod<\/code>\u3002\u800c\u8fd9\u4e2a\u7c7b\u7684 params \u53ea\u6709\u4e09\u4e2a\uff0clookback\u3001neurons\u3001period\u3002<\/p>\n\n\n\n
\u4e09\u5904\u786c\u4f24\uff0c\u90fd\u5728\u4e00\u4e2a\u6587\u4ef6\u91cc\u3002<\/p>\n\n\n\n
\u540c\u4e00\u6279\u626b\u63cf\u91cc\uff0c\u53e6\u5916\u56db\u4e2a\u7b56\u7565\u6587\u4ef6\u90fd\u901a\u8fc7\u4e86\u3002<\/p>\n\n\n\n
\u8fd9\u4e2a\u6587\u4ef6\u8dd1\u4e0d\u8d77\u6765\u8fd8\u6709\u7b2c\u4e8c\u5c42\u539f\u56e0\u3002\u5b83 import \u4e86 rlkit<\/code>\u3002\u8fd9\u4e2a\u5305\u5728 PyPI \u4e0a\u53ea\u6709 0.1 \u548c 0.2.0 \u4e24\u4e2a\u7248\u672c\uff0c\u5168\u90fd\u4f9d\u8d56 tensorflow==1.11.0<\/code>\u3002\u8fd9\u4e2a\u7248\u672c\u7684 tensorflow \u53ea\u63d0\u4f9b\u5230 Python 3.6 \u7684 wheel\u3002<\/p>\n\n\n\n
Qbot \u8981\u6c42 3.8 \u6216 3.9\u3002\u4f9d\u8d56\u94fe\u662f\u6b7b\u9501\u7684\uff0c\u4e0d\u662f\u914d\u7f6e\u95ee\u9898\uff0c\u6539\u4e0d\u51fa\u6765\u3002<\/p>\n\n\n\n
LSTM \u7b56\u7565\u7684\u4fe1\u53f7\u662f\u7528\u672a\u6765\u7b97\u51fa\u6765\u7684<\/strong><\/h2>\n\n\n\n
\u8fd9\u4e2a\u6bd4\u524d\u9762\u7684\u786c\u4f24\u66f4\u503c\u5f97\u8bf4\uff0c\u56e0\u4e3a\u5b83\u4e0d\u62a5\u9519\uff0c\u80fd\u8dd1\u5b8c\uff0c\u8fd8\u80fd\u51fa\u4e00\u5f20\u6f02\u4eae\u7684\u6536\u76ca\u66f2\u7ebf\u3002<\/p>\n\n\n\n
lstm_strategy_bt.py<\/code> \u7684 next()<\/code> \u65b9\u6cd5\u957f\u8fd9\u6837\u3002<\/p>\n\n\n\n
X = self.test_data['X']\ny_pred = self.model.predict(X)\ny_pred = self.scaler.inverse_transform(y_pred)\nif self.dataclose[0] < y_pred[-1]:\n    self.buy(size=100)<\/code><\/pre>\n\n\n\n
\u6bcf\u6b21\u51b3\u7b56\uff0c\u5b83\u90fd\u628a\u6574\u4e2a\u6d4b\u8bd5\u96c6\u4e22\u8fdb\u6a21\u578b\u7b97\u4e00\u904d\uff0c\u7136\u540e\u53d6\u6700\u540e\u4e00\u4e2a\u7b97\u51fa\u6765\u7684\u503c y_pred[-1]\uff0c\u8ddf\u5f53\u524d\u6536\u76d8\u4ef7\u6bd4\u3002<\/p>\n\n\n\n
\u95ee\u9898\u662f\u8fd9\u4e2a y_pred[-1] \u5bf9\u5e94\u7684\u6837\u672c\u7a97\u53e3\uff0c\u662f\u6d4b\u8bd5\u96c6\u7684\u6700\u540e 20 \u5929\u3002<\/p>\n\n\n\n
\u6211\u6309\u8fd9\u6bb5\u4ee3\u7801\u590d\u4e86\u4e00\u904d\u65f6\u95f4\u8f74\u3002<\/p>\n\n\n\n
    \n
  • \u8bad\u7ec3\u6bb5 373 \u6839 K \u7ebf\uff0c\u622a\u81f3 2022 \u5e74 7 \u6708 19 \u65e5<\/li>\n\n\n\n
  • \u51b3\u7b56\u53d1\u751f\u533a\u95f4 94 \u4e2a\u4ea4\u6613\u65e5\uff0c2022 \u5e74 7 \u6708 20 \u65e5\u5230 12 \u6708 6 \u65e5<\/li>\n\n\n\n
  • \u4fe1\u53f7\u5b9e\u9645\u6765\u81ea 2022 \u5e74 11 \u6708 9 \u65e5\u5230 12 \u6708 6 \u65e5\u8fd9\u4e00\u6bb5<\/li>\n<\/ul>\n\n\n\n
    \u4e5f\u5c31\u662f\u8bf4\uff0c7 \u6708 20 \u65e5\u90a3\u5929\u7684\u4e70\u5356\u51b3\u7b56\uff0c\u7528\u7684\u4fe1\u53f7\u6765\u81ea 12 \u6708 6 \u65e5\u6536\u76d8\u7684\u6570\u636e\u3002\u5f80\u524d\u5e73\u5747\u7b97\uff0c\u6bcf\u4e2a\u51b3\u7b56\u70b9\u63d0\u524d\u7528\u4e86\u7ea6 47 \u4e2a\u4ea4\u6613\u65e5\u7684\u672a\u6765\u4fe1\u606f\u3002<\/p>\n\n\n\n
    \"\"
    \u84dd\u8272\u662f\u7b56\u7565\u505a\u51b3\u7b56\u7684\u65f6\u95f4\u6bb5\uff0c\u7ea2\u8272\u662f\u4fe1\u53f7\u5b9e\u9645\u6765\u6e90\u7684\u7a97\u53e3\u3002\u4e24\u4e2a\u533a\u95f4\u51e0\u4e4e\u4e0d\u91cd\u53e0<\/figcaption><\/figure>\n\n\n\n
    \u8fd8\u6709\u4e00\u5904\u3002self.scaler.fit_transform(data)<\/code> \u662f\u5728\u5168\u91cf\u6570\u636e\u4e0a\u62df\u5408\u7684\uff0c\u6d4b\u8bd5\u6bb5\u7684\u4fe1\u606f\u8fdb\u4e86\u5f52\u4e00\u5316\u53c2\u6570\u3002<\/p>\n\n\n\n
    \u8fd9\u7c7b\u4ee3\u7801\u8dd1\u51fa\u6765\u7684\u56de\u6d4b\u7ed3\u679c\u662f\u6ca1\u6709\u610f\u4e49\u7684\u3002\u5b83\u4e0d\u662f\u7b56\u7565\u597d\uff0c\u662f\u5b83\u63d0\u524d\u77e5\u9053\u4e86\u7b54\u6848\u3002\u8fd9\u6761\u6211\u770b\u7b2c\u4e00\u904d\u6ca1\u53cd\u5e94\u8fc7\u6765\uff0c\u7b2c\u4e8c\u904d\u624d\u6709\u70b9\u706b\u5927\u3002<\/p>\n\n\n\n
    \u6211\u7b2c\u4e00\u53cd\u5e94\u662f\u60f3\u53bb issue \u91cc\u7ffb\u6709\u6ca1\u6709\u4eba\u63d0\u8fc7\uff0c\u7ffb\u4e86\u4e00\u5708\u6ca1\u4eba\u63d0\uff0c\u633a\u610f\u5916\u7684\u300268 \u6761\u672a\u5173\u95ed\u7684 issue \u91cc\u6ca1\u6709\u4e00\u6761\u8ddf\u7b56\u7565\u903b\u8f91\u6709\u5173\uff0c\u5168\u662f\u73af\u5883\u95ee\u9898\u3002<\/p>\n\n\n\n
    \u628a\u4e94\u4e2a\u7b56\u7565\u6587\u4ef6\u626b\u5b8c\uff0c\u7ed3\u679c\u662f\u8fd9\u6837\u3002<\/p>\n\n\n\n
    Qbot \u81ea\u5e26\u7684\u7b56\u7565\u6587\u4ef6<\/th>\u9759\u6001\u626b\u63cf\u4e0e\u4f9d\u8d56\u60c5\u51b5<\/th><\/tr>
    rl_strategy_bt.py<\/code> \u5f3a\u5316\u5b66\u4e60<\/td>self.model\u3001self.log \u672a\u5b9a\u4e49\uff0cparams \u91cc\u6ca1\u6709 maperiod\u3002\u4f9d\u8d56 rlkit \u9501\u6b7b tensorflow 1.11\uff0c\u88c5\u4e0d\u4e0a<\/td><\/tr>
    lstm_strategy_bt.py<\/code> LSTM<\/td>\u4ee3\u7801\u80fd\u8dd1\uff0c\u4f46\u4fe1\u53f7\u7528\u4e86\u672a\u6765\u6570\u636e\u3002\u8981 keras \u548c tensorflow<\/td><\/tr>
    sma_cross_strategy_bt.py<\/code> \u5747\u7ebf<\/td>\u901a\u8fc7\u3002\u4f9d\u8d56 tushare\uff0c\u6ca1\u5217\u8fdb requirements<\/td><\/tr>
    boll_strategy_bt.py<\/code> \u5e03\u6797<\/td>\u901a\u8fc7\u3002akshare \u5728 requirements \u91cc\uff0c\u80fd\u88c5<\/td><\/tr>
    multi_strategy_bt.py<\/code> \u591a\u7b56\u7565<\/td>\u901a\u8fc7\u3002\u4f9d\u8d56 tushare\uff0c\u6ca1\u5217\u8fdb requirements<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u6587\u6863\u4e5f\u6709\u95ee\u9898<\/strong><\/h2>\n\n\n\n
    docs\/03-\u667a\u80fd\u7b56\u7565\/model_zoo.md<\/code>\uff0c\u6807\u9898\u5199\u7740 Model Zoo\u3002\u5185\u5bb9\u662f SECOND\u3001SECOND baselines\u30018 \u5f20 GPU \u7684\u663e\u5b58\u5360\u7528\uff0c\u8fd8\u8d34\u4e86\u4e2a mmdetection \u7684 benchmark.py \u94fe\u63a5\u3002<\/p>\n\n\n\n
    \u8fd9\u662f\u4ece\u76ee\u6807\u68c0\u6d4b\u9879\u76ee mmdetection \u590d\u5236\u8fc7\u6765\u7684\uff0c\u8ddf\u91cf\u5316\u4e00\u70b9\u5173\u7cfb\u6ca1\u6709\u3002<\/p>\n\n\n\n
    \u7ecf\u5178\u7b56\u7565\u76ee\u5f55\u91cc\uff0cAlpha\u5bf9\u51b2.md<\/code> \u548c \u6307\u6570\u589e\u5f3a.md<\/code> \u4e24\u4e2a\u6587\u4ef6\uff0c\u5168\u6587\u5404 116 \u5b57\u8282\u3002\u9664\u4e86\u6807\u9898\u548c\u4e00\u53e5”\u4ee5\u4e0b\u90e8\u5206\u5185\u5bb9\u53c2\u8003\u300a\u56e0\u5b50\u6295\u8d44\uff1a\u65b9\u6cd5\u4e0e\u5b9e\u8df5\u300b\u4e00\u4e66”\uff0c\u5c31\u53ea\u5269\u4e00\u4e2a”1. \u539f\u7406”\u548c\u4e00\u884c\u5206\u9694\u7ebf\u3002<\/p>\n\n\n\n
    \u539f\u7406\u6ca1\u4e86\u3002<\/p>\n\n\n\n
    \u8fd8\u6709\u51e0\u4e2a\u6570\u5b57<\/strong><\/h2>\n\n\n\n
    \u6700\u540e push \u662f 2026 \u5e74 3 \u6708 11 \u65e5\uff0c\u4e00\u4e2a autofix \u63d0\u4ea4\uff0c\u4fee\u7684\u662f wxMenuItem \u7684\u95ee\u9898\u3002\u5230\u4eca\u5929\u4e03\u4e2a\u6708\u3002<\/p>\n\n\n\n
    \u8d21\u732e\u8005\u5217\u8868\u62c9\u4e0b\u6765\u53ea\u6709 3 \u4e2a\u4eba\u3002\u4f5c\u8005 Charmve \u63d0\u4ea4 149 \u6b21\uff0c\u53e6\u4e24\u4e2a\u5404 1 \u6b21\u3002<\/p>\n\n\n\n
    68 \u4e2a issue \u6302\u7740\u6ca1\u5173\uff0c29 \u4e2a\u5173\u8fc7\u3002\u5173\u6389\u7684\u90a3 29 \u4e2a\u91cc\u6709\u51e0\u4e2a\u662f\u73af\u5883\u4fee\u590d\u3002<\/p>\n\n\n\n
    fork\/star \u662f 14%\u30021.87 \u4e07 star\uff0c2613 fork\u3002<\/p>\n\n\n\n
    \u8fd9\u4e2a\u6bd4\u4f8b\u8bf4\u660e\u4e00\u4ef6\u4e8b\u3002\u5927\u591a\u6570\u4eba star \u5b8c\u5c31\u8d70\u4e86\uff0c\u771f\u628a\u4ee3\u7801\u62c9\u4e0b\u6765\u6539\u8fc7\u7684\u662f\u90a3 2613 \u4e2a\u4eba\u3002\u800c\u8fd9 2613 \u4e2a\u4eba\u91cc\uff0c\u7edd\u5927\u591a\u6570\u88c5\u73af\u5883\u8fd9\u4e00\u6b65\u5c31\u5361\u4f4f\u4e86\u3002<\/p>\n\n\n\n
    \u90a3\u5b83\u5230\u5e95\u80fd\u7528\u5728\u54ea<\/strong><\/h2>\n\n\n\n
    \u8bf4\u53e5\u516c\u9053\u8bdd\uff0c\u6211\u66f4\u503e\u5411\u628a\u5b83\u5f53\u6210\u4e00\u5957\u9aa8\u67b6\u3002<\/p>\n\n\n\n
    \u8fd9\u5957\u4e1c\u897f\u7684\u4ef7\u503c\u4e0d\u5728\u4e8e\u7b56\u7565\uff0c\u5728\u4e8e\u5b83\u628a\u6570\u636e\u5230\u56de\u6d4b\u8fd9\u4e00\u6bb5\u7684\u8def\u94fa\u597d\u4e86\uff0c\u94fa\u7684\u8fd8\u662f\u56fd\u5185\u80fd\u76f4\u63a5\u7528\u7684\u8def\u3002akshare\u3001efinance\u3001baostock\u3001tushare \u90fd\u662f\u56fd\u5185\u63a5\u53e3\u3002\u6a21\u62df\u4ea4\u6613\u76f4\u63a5\u5bf9\u63a5\u5238\u5546\u5ba2\u6237\u7aef\uff0c\u8d70\u7684\u662f easytrader \u90a3\u5957\u3002GUI \u662f wxPython \u5199\u7684\u684c\u9762\u5e94\u7528\u3002<\/p>\n\n\n\n
    \u5bf9\u4e00\u4e2a\u60f3\u5feb\u901f\u770b\u4e00\u773c\u67d0\u4e2a\u7b56\u7565\u5728 A \u80a1\u4e0a\u4ec0\u4e48\u8868\u73b0\u7684\u4eba\uff0c\u8fd9\u5957\u9aa8\u67b6\u7701\u4e0b\u6765\u7684\u65f6\u95f4\u662f\u771f\u7684\u3002<\/p>\n\n\n\n
    \u5b89\u88c5\u8fd9\u4ef6\u4e8b\uff0c\u6211\u5efa\u8bae\u8fd9\u4e48\u7ed5\u3002<\/p>\n\n\n\n
    Python \u7248\u672c\u9501 3.9\uff0c\u522b\u7528 3.10 \u4ee5\u4e0a\uff08pyfolio \u4f1a\u5d29\uff09\u3002requirements.txt \u91cc matplotlib==3.2.2<\/code> \u6539\u6210 3.7 \u4ee5\u4e0a\u3002git+https:\/\/github.com\/quantopian\/pyfolio<\/code> \u8fd9\u884c\u76f4\u63a5\u5220\u6389\uff0cquantopian 2020 \u5e74\u5c31\u5173\u4e86\uff0c\u4ee3\u7801\u4e0d\u4f1a\u518d\u66f4\u65b0\u3002wxPython \u4ece\u6e90\u7801\u88c5\uff0c\u522b\u7528\u4ed3\u5e93\u91cc\u90a3\u5806\u9884\u7f16\u8bd1 wheel\u3002\u91cc\u9762\u53ea\u6709 cp38 \u548c cp39 \u7684 x86_64 \u7248\u672c\uff0cApple Silicon \u7528\u4e0d\u4e86\u3002<\/p>\n\n\n\n
    \u6570\u636e\u6e90\u522b\u7528\u811a\u672c\u91cc\u9ed8\u8ba4\u7684\u4e1c\u8d22\u63a5\u53e3\uff0c\u6362\u65b0\u6d6a\u3002<\/p>\n\n\n\n
    AI \u7b56\u7565\u90a3\u4e00\u652f\u5148\u522b\u78b0\u3002LSTM \u90a3\u4e2a\u4fe1\u53f7\u662f\u574f\u7684\uff0c\u5f3a\u5316\u5b66\u4e60\u90a3\u4e2a\u4f9d\u8d56\u88c5\u4e0d\u4e0a\u3002\u7ecf\u5178\u7b56\u7565\u91cc\u6709\u80fd\u7528\u7684\uff0c\u5e03\u6797\u3001\u5747\u7ebf\u3001\u6d77\u9f9f\u3001\u5c0f\u5e02\u503c\u3001\u591a\u56e0\u5b50\u8fd9\u51e0\u5957\u4ee3\u7801\u903b\u8f91\u662f\u901a\u7684\uff0c\u6539\u6539\u53c2\u6570\u5c31\u80fd\u8dd1\u3002<\/p>\n\n\n\n
    \u6a21\u62df\u4ea4\u6613\u9700\u8981\u88c5\u5238\u5546\u5ba2\u6237\u7aef\uff0c\u8fd8\u8981\u8fde chromedriver\u3002\u8fd9\u6761\u94fe\u8def\u6211\u6ca1\u80fd\u6d4b\u3002<\/p>\n\n\n\n
    \u6700\u540e<\/strong><\/h2>\n\n\n\n
    \u6211\u628a\u8fd9\u5957\u4e1c\u897f\u62c9\u4e0b\u6765\u88c5\u597d\u8dd1\u901a\uff0c\u82b1\u4e86\u5927\u6982\u4e00\u5929\u3002\u6316\u51fa\u6765\u6700\u6709\u4ef7\u503c\u7684\u4e0d\u662f\u5b83\u6709\u591a\u5c11\u529f\u80fd\uff0c\u662f\u5b83\u54ea\u91cc\u574f\u4e86\u3002\u4f9d\u8d56\u7f3a 11 \u4e2a\u5305\u3002RL \u7b56\u7565\u5f15\u7528\u4e86\u4e0d\u5b58\u5728\u7684\u5c5e\u6027\u3002LSTM \u7684\u4fe1\u53f7\u6765\u81ea\u672a\u6765\u3002\u6587\u6863\u6709\u4e24\u4e2a\u7a7a\u58f3\u3002model_zoo \u662f\u4ece\u76ee\u6807\u68c0\u6d4b\u9879\u76ee\u6284\u7684\u3002<\/p>\n\n\n\n
    \u5f00\u6e90\u9879\u76ee\u7684 star \u6570\u53ea\u8bf4\u660e\u6709\u591a\u5c11\u4eba\u70b9\u4e86\u4e2a\u6536\u85cf\u3002\u88c5\u5f97\u4e0a\u3001\u8dd1\u5f97\u901a\u3001\u8dd1\u51fa\u6765\u7684\u6570\u662f\u5bf9\u7684\uff0c\u662f\u4e09\u4ef6\u5b8c\u5168\u4e0d\u540c\u7684\u4e8b\u3002<\/p>\n\n\n\n
    \u5efa\u8bae\u5927\u5bb6\u53ef\u4ee5\u5148\u81ea\u5df1\u7ffb\u7ffb\u5b83\u7684 issue \u533a\uff0c\u524d 30 \u6761\u6309\u8bc4\u8bba\u6570\u6392\uff0c\u88c5\u4e0d\u4e0a\u7684\u4eba\u8e29\u7684\u5751\u57fa\u672c\u90fd\u5728\u90a3\u91cc\u9762\u3002\u4ed3\u5e93\u5730\u5740\u5728\u6587\u672b\u3002<\/p>\n\n\n\n
    \u53c2\u8003\u8d44\u6599<\/strong><\/h2>\n\n\n\n