{"id":1588,"date":"2024-11-02T07:11:00","date_gmt":"2024-11-01T23:11:00","guid":{"rendered":"https:\/\/blog.laoyulaoyu.top\/?p=1588"},"modified":"2024-10-08T20:14:16","modified_gmt":"2024-10-08T12:14:16","slug":"%e5%b9%b2%e8%b4%a7%ef%bc%81zipline-%e4%b8%8e-pyfolio-%e6%90%ba%e6%89%8b%e7%8e%a9%e8%bd%ac%e7%ae%97%e6%b3%95%e4%ba%a4%e6%98%93%e5%92%8c%e5%9b%9e%e6%ba%af%e6%b5%8b%e8%af%95","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2024\/11\/02\/%e5%b9%b2%e8%b4%a7%ef%bc%81zipline-%e4%b8%8e-pyfolio-%e6%90%ba%e6%89%8b%e7%8e%a9%e8%bd%ac%e7%ae%97%e6%b3%95%e4%ba%a4%e6%98%93%e5%92%8c%e5%9b%9e%e6%ba%af%e6%b5%8b%e8%af%95\/","title":{"rendered":"\u5e72\u8d27\uff01Zipline \u4e0e Pyfolio \u643a\u624b\u73a9\u8f6c\u7b97\u6cd5\u4ea4\u6613\u548c\u56de\u6eaf\u6d4b\u8bd5"},"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
\u5199\u5728\u524d\u9762\u7684\u8bdd\uff1a<\/strong>\u672c\u6587\u4ecb\u7ecd\u4e86\u4f7f\u7528 Python \u5e93 Zipline \u548c Pyfolio \u8fdb\u884c\u7b97\u6cd5\u4ea4\u6613\u7b56\u7565\u7684\u5f00\u53d1\u3001\u56de\u6eaf\u6d4b\u8bd5\u3001\u6027\u80fd\u5206\u6790\u548c\u53ef\u89c6\u5316\u7684\u8fc7\u7a0b<\/mark>\u3002\u6211\u4f1a\u5148\u8bbe\u56de\u6eaf\u6d4b\u8bd5\u73af\u5883\uff0c\u518d\u8bb2\u5747\u503c\u56de\u5f52\u7b56\u7565\u5b9e\u73b0\u3002\u7136\u540e\u7528 Pyfolio \u5206\u6790\u7ee9\u6548\uff0c\u7528 Matplotlib \u548c Seaborn \u53ef\u89c6\u5316\u6570\u636e\u3002\u6700\u540e\u901a\u8fc7\u6848\u4f8b\u6f14\u793a\u8bc4\u4f30\u548c\u4f18\u5316\u7b56\u7565\u3002<\/pre>\n<\/blockquote>\n\n\n\n
\u7b97\u6cd5\u4ea4\u6613\u5f7b\u5e95\u6539\u53d8\u4e86\u91d1\u878d\u7b56\u7565\u7684\u5f00\u53d1\u548c\u6267\u884c\u65b9\u5f0f\u3002Zipline \u662f\u4e00\u4e2a\u7528\u4e8e\u56de\u6eaf\u6d4b\u8bd5\u548c\u5206\u6790\u4ea4\u6613\u7b97\u6cd5\u7684\u5f3a\u5927 Python \u5e93\uff0cPyfolio \u662f\u4e00\u4e2a\u7528\u4e8e\u6027\u80fd\u548c\u98ce\u9669\u5206\u6790\u7684\u5de5\u5177\u3002\u5229\u7528\u8fd9\u4e9b\u5e93\uff0c\u6211\u4eec\u53ef\u4ee5\u8bbe\u8ba1\u548c\u6d4b\u8bd5\u5747\u503c\u56de\u5f52\u7b56\u7565\uff0c\u5206\u6790\u5173\u952e\u6027\u80fd\u6307\u6807\uff0c\u5e76\u5c06\u56de\u62a5\u3001\u6ce2\u52a8\u7387\u3001\u7f29\u6c34\u548c\u6295\u8d44\u7ec4\u5408\u98ce\u9669\u7b49\u5173\u952e\u65b9\u9762\u53ef\u89c6\u5316\u3002\u901a\u8fc7\u8fd9\u4e00\u8fc7\u7a0b\uff0c\u6211\u4eec\u53ef\u4ee5\u6df1\u5165\u4e86\u89e3\u7b56\u7565\u7684\u6709\u6548\u6027\uff0c\u4f18\u5316\u51b3\u7b56\uff0c\u5e76\u66f4\u597d\u5730\u7406\u89e3\u91d1\u878d\u5e02\u573a\u884c\u4e3a\u3002<\/p>\n\n\n\n
import sys\nfrom pathlib import Path\nimport pandas as pd\nfrom pytz import UTC\nfrom zipline import run_algorithm\nfrom zipline.api import (attach_pipeline, date_rules, time_rules, order_target_percent,\n                         pipeline_output, record, schedule_function, get_open_orders, calendars,\n                         set_commission, set_slippage)\nfrom zipline.finance import commission, slippage\nfrom zipline.pipeline import Pipeline, CustomFactor\nfrom zipline.pipeline.factors import Returns, AverageDollarVolume\nimport logbook\nimport matplotlib.pyplot as plt\nimport seaborn<\/code><\/code><\/pre>\n\n\n\n
from pyfolio.utils import extract_rets_pos_txn_from_zipline<\/code><\/code><\/pre>\n\n\n\n
zipline \u56de\u6eaf\u6d4b\u8bd5\u5e93\u7528\u4e8e\u5efa\u7acb\u8fd0\u884c\u7b97\u6cd5\u4ea4\u6613\u7b56\u7565\u7684\u73af\u5883\u3002\u9996\u5148\uff0c\u5b83\u9700\u8981\u5bfc\u5165\u4e00\u4e9b\u57fa\u672c\u6a21\u5757\uff0c\u5982\u7528\u4e8e\u7cfb\u7edf\u64cd\u4f5c\u7684 sys \u548c pathlib\u3001\u7528\u4e8e\u91d1\u878d\u7684 pandas \u548c\u7528\u4e8e\u4ee5 UTC \u5904\u7406\u4e00\u81f4\u65f6\u95f4\u6233\u7684 pytz\u3002\u6b64\u5916\uff0cZipline \u8fd8\u7528\u4e8e\u521b\u5efa\u548c\u7ba1\u7406\u4ea4\u6613\u7b97\u6cd5\uff0c\u56e0\u6b64\u5b83\u4e5f\u5bfc\u5165\u4e86\u8fd9\u4e9b\u7ec4\u4ef6\u3002\u5176\u4e2d\u7684\u5173\u952e\u51fd\u6570\u5305\u62ec run_algorithm\u3001attach_pipeline\u3001pipeline_output\u3001order_target_percent\u3001track performance over time\u3001schedule_function \u548c get_open_orders\u3002\u8fd9\u4e9b\u51fd\u6570\u5bf9\u5728\u7ed9\u5b9a\u65e5\u671f\u8303\u56f4\u5185\u6267\u884c\u4ea4\u6613\u7b56\u7565\u3001attach_pipeline \u548c pipeline_output \u90fd\u5f88\u6709\u7528\u3002<\/p>\n\n\n\n
\u6b64\u5916\uff0c\u4ee3\u7801\u8fd8\u4ece Zipline \u7684\u8d22\u52a1\u6a21\u5757\u4e2d\u5bfc\u5165\u4f63\u91d1\u548c\u6ed1\u70b9\u8bbe\u7f6e\uff0c\u4ee5\u6a21\u62df\u771f\u5b9e\u7684\u4ea4\u6613\u6761\u4ef6\uff0c\u5176\u4e2d\u4f63\u91d1\u4ee3\u8868\u4ea4\u6613\u8fc7\u7a0b\u4e2d\u7684\u8d39\u7528\uff0c\u800c\u6ed1\u70b9\u5219\u4ee3\u8868\u4ef7\u683c\u5dee\u5f02\u3002\u901a\u8fc7\u4f7f\u7528 CustomFactor\uff0c\u53ef\u4ee5\u5b9a\u5236\u6295\u8d44\u7b56\u7565\u3002\u6b64\u5916\uff0c\u8fd8\u5bfc\u5165\u4e86\u7528\u4e8e\u6d3b\u52a8\u8bb0\u5f55\u7684 logbook\u3001\u7528\u4e8e\u6570\u636e\u53ef\u89c6\u5316\u7684 seaborn \u4ee5\u53ca\u7528\u4e8e\u6267\u884c\u540e\u5206\u6790\u7684 Pyfolio \u5de5\u5177\u3002\u57fa\u672c\u4e0a\uff0c\u8fd9\u4e9b\u4ee3\u7801\u4e3a\u6784\u5efa\u3001\u6267\u884c\u548c\u5206\u6790\u7b97\u6cd5\u4ea4\u6613\u7b56\u7565\u5960\u5b9a\u4e86\u57fa\u7840\u3002<\/p>\n\n\n\n
sns.set_style('darkgrid')<\/code><\/code><\/pre>\n\n\n\n
Set_style \u6307\u5b9a\u7ed8\u56fe\u7684\u89c6\u89c9\u5916\u89c2\u3002\u8be5\u884c\u4f7f\u7528 Seaborn \u5e93\uff0c\u8fd9\u662f\u4e00\u4e2a\u6d41\u884c\u7684 Python \u6570\u636e\u53ef\u89c6\u5316\u5e93\u3002Seaborn \u7684 darkgrid \u53c2\u6570\u521b\u5efa\u4e86\u4e00\u79cd\u5177\u6709\u6df1\u8272\u80cc\u666f\u548c\u7f51\u683c\u8986\u76d6\u7684\u6837\u5f0f\uff0c\u53ef\u589e\u5f3a\u7ed8\u56fe\u7684\u53ef\u8bfb\u6027\uff0c\u8ba9\u60a8\u66f4\u5bb9\u6613\u89e3\u91ca\u6570\u636e\u3002\u6df1\u8272\u80cc\u666f\u8ba9\u56fe\u8868\u989c\u8272\u66f4\u5bb9\u6613\u770b\u6e05\u3002\u5728\u521b\u5efa\u53ef\u89c6\u5316\u4e4b\u524d\uff0c\u5b83\u5c31\u4f1a\u4e3a\u4f60\u7684\u53ef\u89c6\u5316\u521b\u5efa\u4e00\u79cd\u98ce\u683c\uff0c\u8ba9\u4f60\u7684\u53ef\u89c6\u5316\u770b\u8d77\u6765\u66f4\u7cbe\u81f4\u3001\u66f4\u4e13\u4e1a\u3002<\/p>\n\n\n\n
# setup stdout logging\nzipline_logging = logbook.NestedSetup([\n    logbook.NullHandler(level=logbook.DEBUG),\n    logbook.StreamHandler(sys.stdout, level=logbook.INFO),\n    logbook.StreamHandler(sys.stderr, level=logbook.ERROR),\n])\nzipline_logging.push_application()<\/code><\/code><\/pre>\n\n\n\n
\u8be5\u4ee3\u7801\u6bb5\u4f7f\u7528\u65e5\u5fd7\u5e93\u914d\u7f6e\u4e86\u591a\u4e2a\u65e5\u5fd7\u5904\u7406\u7a0b\u5e8f\u3002NullHandler \u4f4d\u4e8e DEBUG \u7ea7\u522b\uff0c\u56e0\u6b64\u4f1a\u5ffd\u7565\u65e5\u5fd7\u4fe1\u606f\uff0c\u4e5f\u4e0d\u4f1a\u5bf9\u5904\u7406\u7a0b\u5e8f\u4e22\u5931\u53d1\u51fa\u8b66\u544a\u3002\u7136\u540e\u521b\u5efa\u4e24\u4e2a StreamHandler \u5b9e\u4f8b\uff1a\u4e00\u4e2a\u7528\u4e8e\u5411\u6807\u51c6\u8f93\u51fa\u53d1\u9001 INFO \u6d88\u606f\uff0c\u53e6\u4e00\u4e2a\u7528\u4e8e\u5728\u51fa\u73b0\u4e25\u91cd\u95ee\u9898\u65f6\u5411\u6807\u51c6\u9519\u8bef\u53d1\u9001 ERROR \u6d88\u606f\u3002\u7136\u540e\u8c03\u7528 push_application \u6fc0\u6d3b\u65e5\u5fd7\u8bb0\u5f55\u914d\u7f6e\uff0c\u8fd9\u6837\u5c31\u53ef\u4ee5\u6839\u636e\u4e25\u91cd\u7a0b\u5ea6\u7ba1\u7406\u6d88\u606f\u4e86\u3002<\/p>\n\n\n\n
# Settings\nMONTH = 21\nYEAR = 12 * MONTH\nN_LONGS = 200\nN_SHORTS = 0\nVOL_SCREEN = 1000<\/code><\/code><\/pre>\n\n\n\n
\u4e3a\u4e86\u4fbf\u4e8e\u4ee5\u540e\u4f7f\u7528\uff0c\u6211\u5c06 MONTH \u8bbe\u7f6e\u4e3a 21\uff0c\u8868\u793a\u4ee5\u6708\u4e3a\u5355\u4f4d\u7684\u65f6\u95f4\u6bb5\u3002\u5e74\u7684\u8ba1\u7b97\u65b9\u6cd5\u662f MONTH \u4e58\u4ee5 12\uff0c\u7ed3\u679c\u662f 252 \u4e2a\u6708\u3002\u5728\u4ea4\u6613\u7b56\u7565\u4e2d\uff0cN_LONGS \u8868\u793a\u6709\u591a\u5c11\u591a\u5934\u5934\u5bf8\uff0c\u800c N_SHORTS \u8868\u793a\u6ca1\u6709\u7a7a\u5934\u5934\u5bf8\uff0c\u8fd9\u8868\u660e\u4e70\u5165\u662f\u4f18\u5148\u4e8b\u9879\u3002VOLSCREEN \u8bbe\u7f6e\u4e3a 1000\uff0c\u786e\u4fdd\u53ea\u8003\u8651\u4ea4\u6613\u91cf\u9ad8\u4e8e\u8be5\u9608\u503c\u7684\u8d44\u4ea7\u3002\u8fd9\u4e9b\u5e38\u91cf\u4e3a\u63a5\u4e0b\u6765\u7684\u6b65\u9aa4\u8bbe\u7f6e\u4e86\u53c2\u6570\u3002<\/p>\n\n\n\n
start = pd.Timestamp('2010-01-01', tz=UTC)\nend = pd.Timestamp('2018-01-01', tz=UTC)\ncapital_base = 1e7<\/code><\/code><\/pre>\n\n\n\n
\u4f7f\u7528Panda\u2019s \u65f6\u95f4\u6233\u548c capital_base \u521d\u59cb\u5316\u4e86\u4e24\u4e2a\u53d8\u91cf\u3002Start\u548cend\u88ab\u5b9a\u4e49\u4e3a\u65f6\u95f4\u6233\uff0c\u8868\u793a\u4e00\u4e2a\u65f6\u95f4\u6bb5\u7684\u5f00\u59cb\u548c\u7ed3\u675f\u3002\u4f7f\u7528 pd.Timestamp\uff0c\u6211\u4eec\u7ed9\u51fa\u4e86\u4e00\u4e2a\u5b57\u7b26\u4e32\u65e5\u671f\uff0c\u5e76\u5c06\u65f6\u533a\u8bbe\u7f6e\u4e3a UTC\uff0c\u5f00\u59cb\u548c\u7ed3\u675f\u65f6\u95f4\u8bbe\u7f6e\u4e3a 2010 \u5e74 1 \u6708 1 \u65e5\u3002\u5728\u8d22\u52a1\u5206\u6790\u6216\u5efa\u6a21\u4e2d\uff0ccapital_base \u53ef\u4ee5\u4f5c\u4e3a\u521d\u59cb\u8d44\u672c\u989d\u3002\u65f6\u95f4\u8303\u56f4\u548c\u8d77\u70b9\u5728\u6b64\u8bbe\u7f6e\u3002<\/p>\n\n\n\n
class MeanReversion(CustomFactor):\n    \"\"\"Compute ratio of latest monthly return to 12m average,\n       normalized by std dev of monthly returns\"\"\"\n    inputs = [Returns(window_length=MONTH)]\n    window_length = YEAR<\/code><\/code><\/pre>\n\n\n\n
    def compute(self, today, assets, out, monthly_returns):\n        df = pd.DataFrame(monthly_returns)\n        out[:] = df.iloc[-1].sub(df.mean()).div(df.std())<\/code><\/code><\/pre>\n\n\n\n
MeanReversion \u901a\u8fc7\u6bd4\u8f83\u6700\u8fd1\u4e00\u4e2a\u6708\u7684\u6536\u76ca\u7387\u548c\u8fc7\u53bb\u4e00\u5e74\u4e2d\u5e73\u5747\u4e00\u4e2a\u6708\u7684\u6536\u76ca\u7387\u6765\u8bc4\u4f30\u5747\u503c\u56de\u5f52\uff0c\u5e76\u7528\u6807\u51c6\u5dee\u5bf9\u5dee\u503c\u8fdb\u884c\u5f52\u4e00\u5316\u5904\u7406\u3002\u8f93\u5165\u5c5e\u6027\u6307\u5b9a\u5206\u6790\u4ee5\u6708\u5ea6\u6536\u76ca\u7387\u4e3a\u91cd\u70b9\uff0c\u800c\u6536\u76ca\u7387\u51fd\u6570\u5219\u83b7\u53d6\u4e0a\u4e2a\u6708\u7684\u6570\u636e\u3002\u901a\u8fc7\u5c06\u7a97\u53e3\u957f\u5ea6\u8bbe\u7f6e\u4e3a\u4e00\u5e74\uff0c\u6240\u6709\u8ba1\u7b97\u5747\u4f7f\u7528\u5e74\u5ea6\u6570\u636e\u3002<\/p>\n\n\n\n
\u53c2\u6570\u5305\u62ec\uff1a\u4eca\u5929\u662f\u5f53\u524d\u65e5\u671f\uff0cassets \u662f\u8981\u5206\u6790\u7684\u8d44\u4ea7\uff0cout \u662f\u7ed3\u679c\u7684\u5b58\u50a8\u4f4d\u7f6e\uff0cmonthon_returns \u662f\u53bb\u5e74\u7684\u6570\u636e\u3002\u60a8\u53ef\u4ee5\u5728 pandas DataFrames \u4e2d\u64cd\u4f5c\u6708\u56de\u62a5\u7387\u3002\u53ea\u9700\u51cf\u53bb\u4e4b\u524d\u6708\u5ea6\u56de\u62a5\u7684\u5e73\u5747\u503c\uff0c\u9664\u4ee5\u6807\u51c6\u5dee\uff0c\u7136\u540e\u53d6\u6700\u8fd1\u7684\u6708\u5ea6\u56de\u62a5\u3002\u901a\u8fc7\u5f52\u4e00\u5316\u5904\u7406\uff0c\u4f60\u53ef\u4ee5\u770b\u5230\u6700\u65b0\u6536\u76ca\u7387\u4e0e\u5e73\u5747\u503c\u7684\u504f\u5dee\u6709\u591a\u5927\u3002\u6839\u636e\u524d\u4e00\u5e74\u7684\u8868\u73b0\uff0c\u8f93\u51fa\u6570\u7ec4\u4f1a\u5f97\u5230\u4e00\u4e2a\u5206\u6570\uff0c\u8868\u660e\u6700\u65b0\u6536\u76ca\u7387\u79bb\u7fa4\u7684\u7a0b\u5ea6\u3002<\/p>\n\n\n\n
def compute_factors():\n    \"\"\"Create factor pipeline incl. mean reversion,\n        filtered by 30d Dollar Volume; capture factor ranks\"\"\"\n    mean_reversion = MeanReversion()\n    dollar_volume = AverageDollarVolume(window_length=30)\n    return Pipeline(columns={'longs'  : mean_reversion.bottom(N_LONGS),\n                             'shorts' : mean_reversion.top(N_SHORTS),\n                             'ranking': mean_reversion.rank(ascending=False)},\n                    screen=dollar_volume.top(VOL_SCREEN))<\/code><\/code><\/pre>\n\n\n\n
\u91d1\u878d\u56e0\u7d20\u7ba1\u9053\u901a\u8fc7\u5747\u503c\u56de\u590d\u6cd5\u6765\u627e\u4ea4\u6613\u673a\u4f1a\u3002\u7528 MeanReversion \u5bf9\u8c61\uff0c\u80fd\u6839\u636e\u80a1\u4ef7\u8d70\u52bf\u548c\u56de\u843d\u5230\u5e73\u5747\u4ef7\u7684\u53ef\u80fd\u6027\u7ed9\u80a1\u7968\u6253\u5206\u3002\u4e3a\u4e86\u53ea\u5206\u6790\u6d41\u52a8\u6027\u597d\u7684\u80a1\u7968\uff0c\u4f1a\u5efa\u4e2a\u5e73\u5747\u65e5\u4ea4\u6613\u91cf\u7684\u5b9e\u4f8b\uff0c\u65f6\u95f4\u7a97\u53e3\u662f 30 \u5929\u3002<\/p>\n\n\n\n
\u7ba1\u9053\u7531\u4e09\u4e2a\u4e3b\u8981\u680f\u76ee\u7ec4\u6210\u3002\u591a\u5934\u680f\u6839\u636e\u5747\u503c\u56de\u5f52\u6807\u51c6\u9009\u62e9\u8868\u73b0\u4e0d\u4f73\u7684\u80a1\u7968\uff0c\u53c2\u6570\u4e3a\u591a\u5934\u5934\u5bf8\u7684\u6570\u91cf\u3002\u7ee9\u4f18\u80a1\u5728\u77ed\u7ebf\u680f\u4e2d\u901a\u8fc7\u4f7f\u7528\u76f8\u540c\u7684\u6807\u51c6\u6765\u8bc6\u522b\uff0c\u53c2\u6570\u4e3a\u77ed\u7ebf\u5934\u5bf8\u7684\u6570\u91cf\u3002\u5728\u6392\u540d\u680f\u4e2d\uff0c\u80a1\u7968\u6309\u5747\u503c\u56de\u590d\u903b\u8f91\u4ece\u9ad8\u5230\u4f4e\u6392\u5217\u3002<\/p>\n\n\n\n
\u4f5c\u4e3a\u7ba1\u9053\u7684\u4e00\u90e8\u5206\uff0c\u53ea\u6709\u7b26\u5408\u4ea4\u6613\u91cf\u6807\u51c6\u7684\u80a1\u7968\u624d\u4f1a\u88ab\u7eb3\u5165\u5176\u4e2d\uff0c\u8fd9\u610f\u5473\u7740\u8fd9\u4e9b\u80a1\u7968\u8fd1\u671f\u4ea4\u6613\u60c5\u51b5\u826f\u597d\uff0c\u66f4\u6709\u53ef\u80fd\u53d6\u5f97\u6210\u529f\u3002\u901a\u8fc7\u5747\u503c\u56de\u590d\u5206\u6790\uff0c\u8be5\u529f\u80fd\u53ef\u8bc6\u522b\u591a\u5934\u548c\u7a7a\u5934\u5934\u5bf8\uff0c\u540c\u65f6\u901a\u8fc7\u7f8e\u5143\u6210\u4ea4\u91cf\u8fc7\u6ee4\u5668\u786e\u4fdd\u8db3\u591f\u7684\u6d41\u52a8\u6027\u3002<\/p>\n\n\n\n
def before_trading_start(context, data):\n    \"\"\"Run factor pipeline\"\"\"\n    context.factor_data = pipeline_output('factor_pipeline')\n    record(factor_data=context.factor_data.ranking)\n    assets = context.factor_data.index\n    record(prices=data.current(assets, 'price'))<\/code><\/code><\/pre>\n\n\n\n
\u5728\u4ea4\u6613\u5f00\u59cb\u65f6\uff0cbefore_trading_start \u51fd\u6570\u4f1a\u8fd0\u884c\u4e00\u4e2a\u56e0\u5b50\u6d41\u6c34\u7ebf\uff0c\u7528\u4e8e\u6839\u636e\u91cf\u5316\u91d1\u878d\u4e2d\u7684\u67d0\u4e9b\u6807\u51c6\u5bf9\u80a1\u7968\u8fdb\u884c\u8bc4\u4f30\u3002\u8c03\u7528\u8be5\u51fd\u6570\u65f6\uff0c\u5b83\u4f1a\u4f7f\u7528 pipeline_output \u51fd\u6570\u68c0\u7d22\u56e0\u5b50\u6d41\u6c34\u7ebf\u7684\u8f93\u51fa\uff0c\u5e76\u5c06\u5176\u5b58\u50a8\u5230\u5305\u542b\u5927\u91cf\u8d44\u4ea7\u4fe1\u606f\u7684 context.factor_data \u4e2d\u3002\u4e3a\u4e86\u5206\u6790\u968f\u65f6\u95f4\u53d8\u5316\u7684\u6027\u80fd\uff0c\u8be5\u51fd\u6570\u4f7f\u7528 record \u51fd\u6570\u8bb0\u5f55\u6392\u540d\u3002assets \u53d8\u91cf\u6839\u636e context.factor_data \u4e2d\u7684\u7d22\u5f15\u63d0\u4f9b\u76f8\u5173\u8d44\u4ea7\u5217\u8868\u3002\u5b83\u8fd8\u4f7f\u7528 data.current \u65b9\u6cd5\u83b7\u53d6\u5e76\u8bb0\u5f55\u8fd9\u4e9b\u8d44\u4ea7\u7684\u5f53\u524d\u4ef7\u683c\uff0c\u8fd9\u5bf9\u4e86\u89e3\u5e02\u573a\u80cc\u666f\u975e\u5e38\u91cd\u8981\u3002\u8fd9\u662f\u4ea4\u6613\u6d3b\u52a8\u524d\u8fdb\u884c\u51c6\u5907\u7684\u91cd\u8981\u529f\u80fd\uff0c\u53ef\u4ee5\u6839\u636e\u5e02\u573a\u60c5\u51b5\u505a\u51fa\u6570\u636e\u9a71\u52a8\u7684\u51b3\u7b56\u3002<\/p>\n\n\n\n
def rebalance(context, data):\n    \"\"\"Compute long, short and obsolete holdings; place trade orders\"\"\"\n    factor_data = context.factor_data\n    assets = factor_data.index<\/code><\/code><\/pre>\n\n\n\n
    longs = assets[factor_data.longs]\n    shorts = assets[factor_data.shorts]\n    divest = context.portfolio.positions.keys() - longs.union(shorts)<\/code><\/code><\/pre>\n\n\n\n
exec_trades(data, assets=divest, target_percent=0)\n    exec_trades(data, assets=longs, target_percent=1 \/ N_LONGS if N_LONGS else 0)\n    exec_trades(data, assets=shorts, target_percent=-1 \/ N_SHORTS if N_SHORTS else 0)<\/code><\/code><\/pre>\n\n\n\n
\u518d\u5e73\u8861\u529f\u80fd\u4e2d\uff0c\u8fc7\u65f6\u7684\u6301\u4ed3\u4f1a\u88ab\u6e05\u7b97\uff0c\u591a\u5934\u548c\u7a7a\u5934\u5934\u5bf8\u4f1a\u6839\u636e\u60a8\u63d0\u4f9b\u7684\u56e0\u5b50\u8fdb\u884c\u518d\u5e73\u8861\u3002\u56e0\u5b50\u6570\u636e\u662f\u7b2c\u4e00\u6b65\uff0c\u5176\u4e2d\u5305\u542b\u5404\u79cd\u8d44\u4ea7\u548c\u4ea4\u6613\u4fe1\u53f7\u7684\u4fe1\u606f\u3002\u8d44\u4ea7\u662f\u6307\u8fd9\u4e9b\u6570\u636e\u4e2d\u5305\u542b\u5728\u8d44\u4ea7\u53d8\u91cf\u4e2d\u7684\u8d44\u4ea7\u3002<\/p>\n\n\n\n
\u8be5\u51fd\u6570\u5b9a\u4e49\u4e86\u4e09\u7ec4\u8d44\u4ea7\uff1a\u591a\u5934\uff0c\u5373\u591a\u5934\u5934\u5bf8\uff1b\u7a7a\u5934\uff0c\u5373\u7a7a\u5934\u5934\u5bf8\uff1b\u5265\u79bb\uff0c\u5373\u5f53\u524d\u6295\u8d44\u7ec4\u5408\u4e2d\u6ca1\u6709\u591a\u5934\u6216\u7a7a\u5934\u5934\u5bf8\u7684\u8d44\u4ea7\uff0c\u8ba1\u7b97\u65b9\u6cd5\u662f\u51cf\u53bb\u591a\u5934\u548c\u7a7a\u5934\u8d44\u4ea7\u7684\u603b\u548c\u3002<\/p>\n\n\n\n
\u8c03\u7528\u5265\u79bb\u8d44\u4ea7\u65f6\uff0c\u4f1a\u8c03\u7528\u76ee\u6807\u767e\u5206\u6bd4\u4e3a 0 \u7684 exec_trades\u3002\u6839\u636e\u591a\u5934\u5934\u5bf8\u7684\u6570\u91cf\uff0c\u5b83\u5c06\u6295\u8d44\u5e73\u5747\u5206\u914d\u7ed9\u8fd9\u4e9b\u591a\u5934\u5934\u5bf8\uff0c\u5982\u679c\u6ca1\u6709\u591a\u5934\u5934\u5bf8\uff0c\u5219\u8bbe\u7f6e 0%\u3002\u5229\u7528\u7a7a\u5934\u5934\u5bf8\u7684\u6570\u91cf\uff0c\u786e\u4fdd\u8d44\u91d1\u5e73\u5747\u5206\u914d\u5230\u8d1f\u65b9\u5411\u3002<\/p>\n\n\n\n
\u8fd9\u6837\uff0c\u6295\u8d44\u7ec4\u5408\u5c31\u80fd\u6839\u636e\u6700\u65b0\u7684\u4ea4\u6613\u4fe1\u53f7\u4e0d\u65ad\u4f18\u5316\uff0c\u5934\u5bf8\u4e5f\u80fd\u5f97\u5230\u6709\u6548\u7ba1\u7406\u3002<\/p>\n\n\n\n
def exec_trades(data, assets, target_percent):\n    \"\"\"Place orders for assets using target portfolio percentage\"\"\"\n    for asset in assets:\n        if data.can_trade(asset) and not get_open_orders(asset):\n            order_target_percent(asset, target_percent)<\/code><\/code><\/pre>\n\n\n\n
\u5728 exec_trades \u51fd\u6570\u4e2d\uff0c\u60a8\u9700\u8981\u7ed9\u51fa\u4e09\u4e2a\u53c2\u6570\uff1a\u6570\u636e\u3001\u8d44\u4ea7\u548c target_percent\u3002\u5b83\u4f1a\u6839\u636e\u76ee\u6807\u767e\u5206\u6bd4\u4e3a\u7279\u5b9a\u8d44\u4ea7\u4e0b\u5355\u3002\u8be5\u51fd\u6570\u4f1a\u68c0\u67e5\u6bcf\u79cd\u8d44\u4ea7\u662f\u5426\u5141\u8bb8\u4ea4\u6613\uff0c\u4ee5\u53ca\u8d44\u4ea7\u5217\u8868\u4e2d\u662f\u5426\u6ca1\u6709\u8be5\u8d44\u4ea7\u7684\u672a\u7ed3\u8ba2\u5355\u3002\u4e00\u65e6\u8fd9\u4e24\u4e2a\u6761\u4ef6\u90fd\u6ee1\u8db3\uff0c\u5b83\u5c31\u4f1a\u8c03\u7528 order_target_percent\uff0c\u8bbe\u7f6e\u8d44\u4ea7\u7684\u76ee\u6807\u5206\u914d\u3002\u8fd9\u79cd\u65b9\u6cd5\u53ea\u6709\u5728\u5141\u8bb8\u4e0b\u5355\u548c\u6ca1\u6709\u6302\u5355\u7684\u60c5\u51b5\u4e0b\u624d\u4f1a\u4e0b\u5355\uff0c\u4ece\u800c\u786e\u4fdd\u9ad8\u6548\u7684\u4ea4\u6613\u7ba1\u7406\u3002<\/p>\n\n\n\n
def initialize(context):\n    \"\"\"Setup: register pipeline, schedule rebalancing,\n        and set trading params\"\"\"\n    attach_pipeline(compute_factors(), 'factor_pipeline')\n    schedule_function(rebalance,\n                      date_rules.week_start(),\n                      time_rules.market_open(),\n                      calendar=calendars.US_EQUITIES)<\/code><\/code><\/pre>\n\n\n\n
set_commission(us_equities=commission.PerShare(cost=0.00075, min_trade_cost=.01))\n    set_slippage(us_equities=slippage.VolumeShareSlippage(volume_limit=0.0025, price_impact=0.01))<\/code><\/code><\/pre>\n\n\n\n
\u521d\u59cb\u5316\u51fd\u6570\u4e3a\u4ea4\u6613\u7b97\u6cd5\u8bbe\u7f6e\u73af\u5883\u3002\u5b83\u914d\u7f6e\u5fc5\u8981\u7684\u7ec4\u4ef6\uff0c\u4f7f\u5176\u80fd\u6b63\u5e38\u5de5\u4f5c\u3002\u4f7f\u7528 attach_pipeline \u6ce8\u518c\u7ba1\u9053\uff0c\u6536\u96c6\u6570\u636e\u5e76\u8ba1\u7b97\u5f71\u54cd\u4ea4\u6613\u51b3\u7b56\u7684\u56e0\u7d20\u3002\u518d\u5e73\u8861\u51fd\u6570\u4f1a\u5728\u6bcf\u5468\u5f00\u5e02\u65f6\u8c03\u6574\u6295\u8d44\u7ec4\u5408\u7684\u6301\u6709\u91cf\uff0c\u786e\u4fdd\u4ea4\u6613\u6309\u65f6\u6267\u884c\u3002<\/p>\n\n\n\n
\u6b64\u5916\uff0c\u4f63\u91d1\u662f\u901a\u8fc7 set_commission \u8bbe\u7f6e\u7684\uff0c\u6700\u4f4e\u4ef7\u683c\u4e3a 0.01 \u7f8e\u5143\uff0c\u6bcf\u80a1\u6210\u672c\u4e3a 0.00075 \u7f8e\u5143\u3002\u8fd9\u6837\uff0c\u5373\u4f7f\u4ea4\u6613\u91cf\u4e0d\u5927\uff0c\u4e5f\u6709\u6700\u4f4e\u8d39\u7528\u3002\u901a\u8fc7\u5c06\u6ed1\u70b9\u8bbe\u7f6e\u4e3a 0.0025\uff0c\u5e76\u5047\u8bbe\u6bcf\u7b14\u4ea4\u6613\u5bf9\u4ef7\u683c\u7684\u5f71\u54cd\u4e3a 1%\uff0c\u8be5\u4ea4\u6613\u7b56\u7565\u53ef\u4ee5\u6709\u6548\u5730\u6536\u96c6\u6570\u636e\u3001\u7ba1\u7406\u6301\u80a1\uff0c\u5e76\u5728\u89c4\u5b9a\u7684\u6210\u672c\u548c\u98ce\u9669\u8303\u56f4\u5185\u8fd0\u4f5c\u3002<\/p>\n\n\n\n
backtest = run_algorithm(start=start,\n                         end=end,\n                         initialize=initialize,\n                         before_trading_start=before_trading_start,\n                         capital_base=capital_base)<\/code><\/code><\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\u56de\u6eaf\u6d4b\u8bd5\u4f7f\u7528 Zipline \u7684 run_algorithm \u51fd\u6570\u8fdb\u884c\uff0cZipline \u662f\u4e00\u4e2a\u7b97\u6cd5\u4ea4\u6613\u5e93\u3002\u5f00\u59cb\u548c\u7ed3\u675f\u662f\u7528\u4e8e\u56de\u6eaf\u6d4b\u8bd5\u7684\u53c2\u6570\uff0cinitialize \u7528\u4e8e\u8bbe\u7f6e\u4ea4\u6613\u7b97\u6cd5\uff0cbefore_trading_start \u7528\u4e8e\u8fdb\u884c\u4ea4\u6613\u524d\u8c03\u6574\uff0ccapital_base \u7528\u4e8e\u6307\u5b9a\u7b56\u7565\u7684\u521d\u59cb\u8d44\u91d1\u3002<\/p>\n\n\n\n
\u5728\u8f93\u51fa\u65e5\u5fd7\u4e2d\uff0c\u60a8\u53ef\u4ee5\u770b\u5230\u56de\u6eaf\u6d4b\u8bd5\u7684\u6267\u884c\u60c5\u51b5\uff0c\u4e00\u5f00\u59cb\u4f1a\u51fa\u73b0\u57fa\u51c6\u6570\u636e\u4e0d\u53ef\u7528\u7684\u8b66\u544a\uff0c\u4f46\u540e\u6765\u786e\u8ba4\u6570\u636e\u5df2\u7ecf\u4e0b\u8f7d\u3002\u5728\u56de\u6eaf\u6d4b\u8bd5\u671f\u95f4\uff0c\u6027\u80fd\u6307\u6807\u663e\u793a\u5728\u6807\u6709 INFO \u7684\u65e5\u5fd7\u6761\u76ee\u4e2d\u3002\u5b83\u4eec\u663e\u793a\u4e86\u5404\u79cd\u8d44\u4ea7\u7684\u91d1\u989d\u3001\u6210\u672c\u57fa\u7840\u548c\u73b0\u91d1\u4f59\u989d\u3002<\/p>\n\n\n\n
\u4f8b\u5982\uff0c\u80a1\u7968(3064 [WCN])\u548c\u80a1\u7968(3070 [WEC])\u7684\u4e1a\u7ee9\u8fde\u540c\u5176\u91d1\u989d\u548c\u6210\u672c\u57fa\u7840\u90fd\u8bb0\u5f55\u5728\u6848\u3002\u5728\u4ea4\u6613\u8fc7\u7a0b\u4e2d\uff0c\u73b0\u91d1\u4f59\u989d\u4f1a\u53d1\u751f\u53d8\u5316\uff0c\u53cd\u6620\u4ea4\u6613\u7ed3\u679c\u3002\u6b64\u5916\uff0c\u5206\u5f55\u8fd8\u544a\u8bc9\u4f60\u80a1\u7968\u62c6\u5206\u540e\u7684\u8868\u73b0\uff0c\u8fd9\u5bf9\u4e8e\u5f04\u6e05\u80a1\u7968\u62c6\u5206\u5bf9\u6295\u8d44\u7ec4\u5408\u4ef7\u503c\u7684\u5f71\u54cd\u975e\u5e38\u91cd\u8981\u3002<\/p>\n\n\n\n
\u5728\u56de\u6eaf\u6d4b\u8bd5\u8fc7\u7a0b\u4e2d\uff0c\u65e5\u5fd7\u4f1a\u5b9a\u671f\u663e\u793a\u4ea4\u6613\u7b56\u7565\u5bf9\u4e1a\u7ee9\u7684\u5f71\u54cd\u3002\u7531\u4e8e\u6700\u540e\u7684\u6761\u76ee\uff0c\u56de\u6eaf\u6d4b\u8bd5\u8de8\u8d8a\u4e86 2013 \u5e74\uff0c\u56e0\u6b64\u7528\u6237\u53ef\u4ee5\u8bc4\u4f30\u5176\u4ea4\u6613\u7b56\u7565\u7684\u6709\u6548\u6027\u3001\u8bc4\u4f30\u98ce\u9669\u5e76\u505a\u51fa\u660e\u667a\u7684\u8c03\u6574\u3002<\/p>\n\n\n\n
returns, positions, transactions = extract_rets_pos_txn_from_zipline(backtest)<\/code><\/code><\/pre>\n\n\n\n
\u8fd9\u884c\u4ee3\u7801\u8c03\u7528\u4ee5 backtest \u53d8\u91cf\u4e3a\u53c2\u6570\u7684 extract_rets_pos_txn_from_zipline \u51fd\u6570\u3002\u5b83\u4f7f\u7528 Python \u7b97\u6cd5\u4ea4\u6613\u5e93 Zipline \u83b7\u53d6\u56de\u6eaf\u6d4b\u8bd5\u7684\u5177\u4f53\u4fe1\u606f\u3002\u5176\u4e2d\u6709\u4e09\u4e2a\u6570\u636e\u70b9\uff1a\u6536\u76ca\u3001\u4ed3\u4f4d\u548c\u4ea4\u6613\u3002\u6536\u76ca\u5305\u542b\u7b56\u7565\u6027\u80fd\uff0c\u5982\u5229\u6da6\u6216\u4e8f\u635f\u3002\u4ed3\u4f4d\u663e\u793a\u4f60\u5728\u56de\u6eaf\u6d4b\u8bd5\u671f\u95f4\u6301\u6709\u7684\u8d44\u4ea7\u3002\u4ea4\u6613\u8be6\u7ec6\u8bb0\u5f55\u4e86\u6574\u4e2a\u56de\u6eaf\u6d4b\u8bd5\u671f\u95f4\u6267\u884c\u7684\u4ea4\u6613\uff0c\u5305\u62ec\u4e70\u5165\u548c\u5356\u51fa\u6d3b\u52a8\u3002\u5143\u7ec4\u89e3\u5305\u4f7f\u5f97\u5206\u914d\u8f93\u51fa\u53d8\u5f97\u7b80\u5355\u3002<\/p>\n\n\n\n
with pd.HDFStore('backtests.h5') as store:\n    store.put('backtest', backtest)\n    store.put('returns', returns)\n    store.put('positions', positions)\n    store.put('transactions', transactions)<\/code><\/code><\/pre>\n\n\n\n
\u4f7f\u7528 pandas \u5e93\u5c06\u56de\u6eaf\u6d4b\u8bd5\u8fc7\u7a0b\u4e2d\u7684\u6570\u636e\u7ec4\u4ef6\u5b58\u50a8\u5728 HDF5 \u6587\u4ef6\u4e2d\u3002HDFStore \u4e0a\u4e0b\u6587\u7ba1\u7406\u5668\u786e\u4fdd\u6570\u636e\u5b89\u5168\u65e0\u865e\u3002\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0cput \u65b9\u6cd5\u4f1a\u4fdd\u5b58\u6570\u636e\u96c6\uff0c\u5982\u56de\u6eaf\u6d4b\u8bd5\u3001\u56de\u62a5\u3001\u4ed3\u4f4d\u548c\u4ea4\u6613\uff0c\u8fd9\u4e9b\u6570\u636e\u96c6\u5bf9\u56de\u6eaf\u6d4b\u8bd5\u7ed3\u679c\u975e\u5e38\u91cd\u8981\uff0c\u5e76\u4f7f\u6570\u636e\u5b58\u50a8\u548c\u68c0\u7d22\u66f4\u52a0\u5bb9\u6613\u3002<\/p>\n\n\n\n
\u5728 HDF5 \u4e2d\u5e8f\u5217\u5316\u67d0\u4e9b\u5bf9\u8c61\u7c7b\u578b\u53ef\u80fd\u4f1a\u51fa\u73b0\u95ee\u9898\uff0c\u56e0\u4e3a\u6570\u636e\u53ef\u80fd\u5305\u542b\u65e0\u6cd5\u6620\u5c04\u5230 C \u7c7b\u578b\u7684\u5143\u7d20\u3002\u6709\u5173\u6027\u80fd\u95ee\u9898\u7684\u8b66\u544a\u5f3a\u8c03\u4e86\u786e\u4fdd\u5b58\u50a8\u6570\u636e\u4e0e HDF5 \u517c\u5bb9\u7684\u5fc5\u8981\u6027\uff0c\u5c24\u5176\u662f\u5bf9\u4e8e\u590d\u6742\u7684\u6570\u636e\u7ed3\u6784\u548c\u81ea\u5b9a\u4e49\u5bf9\u8c61\u3002<\/p>\n\n\n\n
fig, axes= plt.subplots(nrows=2, figsize=(14,6))\nreturns.add(1).cumprod().sub(1).plot(ax=axes[0], title='Cumulative Returns')\ntransactions.groupby(transactions.dt.dt.day).txn_dollars.sum().cumsum().plot(ax=axes[1], title='Cumulative Transactions')\nfig.tight_layout();<\/code><\/code><\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\u8be5\u4ee3\u7801\u7247\u6bb5\u53ef\u751f\u6210\u4e00\u4e2a\u5305\u542b\u4e24\u4e2a\u5b50\u56fe\u7684\u56fe\u8868\uff0c\u76f4\u89c2\u663e\u793a\u4f7f\u7528 Zipline \u8fdb\u884c\u5747\u503c\u56de\u5f52\u56de\u6eaf\u6d4b\u8bd5\u6240\u83b7\u5f97\u7684\u8d22\u52a1\u6570\u636e\u3002\u7b2c\u4e00\u4e2a\u5b50\u56fe\u663e\u793a\u968f\u65f6\u95f4\u63a8\u79fb\u7684\u7d2f\u8ba1\u6536\u76ca\uff0c\u7b2c\u4e8c\u4e2a\u5b50\u56fe\u663e\u793a\u968f\u65f6\u95f4\u63a8\u79fb\u7684\u7d2f\u8ba1\u4ea4\u6613\u503c\u3002<\/p>\n\n\n\n
\u5c06\u6536\u76ca\u52a0\u4e00\u5e76\u8ba1\u7b97\u7d2f\u8ba1\u4e58\u79ef\uff0c\u53ef\u4ee5\u8ddf\u8e2a\u7b2c\u4e00\u5e45\u5b50\u56fe\u4e2d\u4ece\u4e00\u5f00\u59cb\u7684\u6295\u8d44\u589e\u957f\u60c5\u51b5\u3002\u5b83\u901a\u5e38\u5448\u4e0a\u5347\u8d8b\u52bf\uff0c\u56e0\u6b64\u6295\u8d44\u7b56\u7565\u662f\u6709\u5229\u53ef\u56fe\u7684\uff0c\u4f46\u4e5f\u65f6\u6709\u6ce2\u52a8\u3002<\/p>\n\n\n\n
\u5b83\u901a\u8fc7\u5c06\u6bcf\u65e5\u4ea4\u6613\u91d1\u989d\u76f8\u52a0\u5e76\u8ba1\u7b97\u7d2f\u8ba1\u91d1\u989d\u6765\u663e\u793a\u4e00\u6bb5\u65f6\u95f4\u5185\u7684\u603b\u4ea4\u6613\u91cf\u3002\u7531\u4e8e\u4ea4\u6613\u6d3b\u52a8\u7684\u6ce2\u52a8\u6027\uff0c\u8be5\u56fe\u8c31\u6bd4\u7d2f\u8ba1\u6536\u76ca\u66f4\u5bb9\u6613\u53d8\u5316\u3002<\/p>\n\n\n\n
\u8fd9\u79cd\u5e03\u5c40\u4fbf\u4e8e\u6bd4\u8f83\u7d2f\u8ba1\u56de\u62a5\u548c\u4ea4\u6613\u6d3b\u52a8\u3002\u901a\u8fc7\u8fd9\u4e9b\u53ef\u89c6\u5316\u56fe\u8868\uff0c\u6211\u4eec\u53ef\u4ee5\u6df1\u5165\u4e86\u89e3\u4ea4\u6613\u7b56\u7565\u7684\u8868\u73b0\u548c\u4ea4\u6613\u52a8\u6001\u3002<\/p>\n\n\n\n
positions.info()<\/code><\/code><\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
Positions.info() \u4f7f\u7528\u7684\u662f Pandas\uff0c\u53ef\u4ee5\u8ba9\u4f60\u5728 Python \u4e2d\u5206\u6790\u548c\u5904\u7406\u6570\u636e\u3002\u8fd9\u662f DataFrame \u7684\u6458\u8981\uff0c\u5728\u672c\u4f8b\u4e2d\uff0c\u5b83\u8ddf\u8e2a\u7684\u662f\u4f7f\u7528\u5747\u503c\u56de\u590d\u7b56\u7565 Zipline \u6846\u67b6\u7684\u56de\u6eaf\u6d4b\u8bd5\u9879\u76ee\u4e2d\u7684\u5934\u5bf8\u3002\u5728\u4e00\u4e2a\u76f8\u5f53\u957f\u7684\u65f6\u95f4\u6bb5\u5185\uff0cDataFrame \u5305\u542b 1717 \u4e2a\u6761\u76ee\uff0c\u6309\u65f6\u95f4\u987a\u5e8f\u6392\u5217\uff0c\u4fbf\u4e8e\u5206\u6790\u3002\u80a1\u7968\u548c\u73b0\u91d1\u8868\u793a\u5b83\u4ee3\u8868\u4e86\u4e00\u6bb5\u65f6\u95f4\u5185\u7684\u8d44\u4ea7\u548c\u6295\u8d44\u4ef7\u503c\uff0c\u91cd\u70b9\u662f\u73b0\u91d1\u3002\u91c7\u7528 float64 \u6570\u636e\u7c7b\u578b\uff0c\u6570\u503c\u4e3a\u6d6e\u70b9\u6570\uff0c\u975e\u5e38\u9002\u5408\u5305\u542b\u5c0f\u6570\u7684\u8d22\u52a1\u6570\u636e\u3002\u8003\u8651\u5230\u5176\u6761\u76ee\u6570\u91cf\uff0c\u5b83\u7684\u5185\u5b58\u4f7f\u7528\u91cf\u76f8\u5f53\u9002\u4e2d\uff0c\u8fd9\u5bf9\u63d0\u9ad8\u6570\u636e\u5904\u7406\u6548\u7387\u5f88\u6709\u5e2e\u52a9\u3002Positions.info() \u53ef\u4ee5\u8ba9\u7528\u6237\u6df1\u5165\u4e86\u89e3 DataFrame \u7684\u7ed3\u6784\u548c\u5185\u5bb9\uff0c\u4ece\u800c\u5224\u65ad\u6570\u636e\u662f\u5426\u53ef\u4ee5\u7528\u4e8e\u8fdb\u4e00\u6b65\u5206\u6790\u6216\u5efa\u6a21\u3002<\/p>\n\n\n\n
transactions.describe()<\/code><\/code><\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\u5728\u5747\u503c\u56de\u5f52\u56de\u6eaf\u6d4b\u8bd5\u9879\u76ee\u4e2d\uff0ctransactions.describe() \u4e3a\u540d\u4e3a transactions \u7684 DataFrame \u751f\u6210\u63cf\u8ff0\u6027\u7edf\u8ba1\u3002\u6bcf\u5217\u6709 54,596 \u4e2a\u6761\u76ee\uff0c\u8868\u660e\u8fd9\u662f\u4e00\u4e2a\u5e9e\u5927\u7684\u6570\u636e\u96c6\uff0c\u5176\u4e2d\u6709\u4e09\u4e2a\u5173\u952e\u5217\uff1a\u91d1\u989d\u3001\u4ef7\u683c\u548c txn_dollars\u3002\u5927\u7ea6\u6709 6.64 \u7b14\u4ea4\u6613\uff0c\u6bcf\u5355\u4f4d\u5e73\u5747\u4ef7\u683c\u4e3a 54.53\uff0c\u5e73\u5747\u91d1\u989d\u4e3a -125.97\uff0c\u8fd9\u8868\u660e\u6570\u636e\u96c6\u4e3b\u8981\u662f\u9500\u552e\u6216\u635f\u5931\u3002<\/p>\n\n\n\n
\u4ece\u6807\u51c6\u504f\u5dee\u503c\u53ef\u4ee5\u770b\u51fa\uff0c\u8be5\u91d1\u989d\u7684\u6807\u51c6\u504f\u5dee\u5f88\u9ad8\uff0c\u4e3a 3164.43\uff0c\u8868\u660e\u5176\u6ce2\u52a8\u5f88\u5927\u3002-19,313\u662f\u6700\u5c0f\u4ea4\u6613\u989d\uff0c\u8868\u660e\u6709\u5f88\u591a\u629b\u552e\u6216\u8d1f\u4ea4\u6613\uff0c\u800c312,268\u662f\u6700\u5927\u4ea4\u6613\u989d\uff0c\u8868\u660e\u6709\u5f88\u591a\u4ea4\u6613\u3002\u5728\u56db\u5206\u4f4d\u6570\u4e2d\uff0c\u91d1\u989d\u7684\u7b2c 25 \u767e\u5206\u4f4d\u6570\u4e3a -798\uff0c\u8fd9\u610f\u5473\u7740\u6709 25% \u7684\u4ea4\u6613\u4f4e\u4e8e\u8be5\u503c\uff1b\u4e2d\u4f4d\u6570\u4e3a 7\uff0c\u8868\u660e\u6709\u4e00\u534a\u7684\u4ea4\u6613\u4f4e\u4e8e\u8be5\u503c\uff1b\u7b2c 75 \u767e\u5206\u4f4d\u6570\u4e3a 790.25\uff0c\u8fd9\u610f\u5473\u7740\u5927\u591a\u6570\u4ea4\u6613\u6ca1\u6709\u90a3\u4e48\u5927\u3002<\/p>\n\n\n\n
\u901a\u8fc7\u8fd9\u79cd\u5206\u6790\uff0c\u6211\u4eec\u53ef\u4ee5\u5bf9\u4ea4\u6613\u6570\u636e\u6709\u4e00\u4e2a\u6e05\u6670\u7684\u6982\u89c8\uff0c\u63ed\u793a\u6a21\u5f0f\u548c\u5f02\u5e38\u503c\uff0c\u5e2e\u52a9\u6211\u4eec\u4e86\u89e3\u4ea4\u6613\u7b56\u7565\u7684\u6027\u80fd\u548c\u884c\u4e3a\u3002<\/p>\n\n\n\n
from pathlib import Path\nimport warnings\nimport re\nimport numpy as np\nimport pandas as pd\nimport seaborn as sns\nimport matplotlib.pyplot as plt<\/code><\/code><\/pre>\n\n\n\n
pathlib \u4e2d\u7684 Path \u7b80\u5316\u4e86\u6587\u4ef6\u8def\u5f84\u64cd\u4f5c\u3002\u8be5\u4ee3\u7801\u7247\u6bb5\u5bfc\u5165\u4e86\u7528\u4e8e\u6570\u636e\u5904\u7406\u3001\u53ef\u89c6\u5316\u548c\u6587\u4ef6\u5904\u7406\u7684 Python \u5e93\u548c\u6a21\u5757\u3002\u4f7f\u7528 warnings \u6a21\u5757\uff0c\u53ef\u4ee5\u63a7\u5236\u8b66\u544a\u4fe1\u606f\u7684\u663e\u793a\u65b9\u5f0f\uff0c\u4ece\u800c\u907f\u514d\u5d29\u6e83\u3002\u4f7f\u7528 re \u6a21\u5757\uff0c\u53ef\u4ee5\u641c\u7d22\u5b57\u7b26\u4e32\uff0c\u5e76\u4f7f\u7528\u6b63\u5219\u8868\u8fbe\u5f0f\u5339\u914d\u6a21\u5f0f\u3002\u6709\u4e86 Numpy\uff0c\u4f60\u53ef\u4ee5\u8fd0\u884c\u6570\u7ec4\u5e76\u4f7f\u7528\u6570\u5b66\u51fd\u6570\u3002Pandas \u662f\u5206\u6790\u7ed3\u6784\u5316\u6570\u636e\u7684\u5b8c\u7f8e\u5de5\u5177\uff0cDataFrames \u53ef\u8ba9\u60a8\u8f7b\u677e\u5904\u7406\u548c\u6e05\u7406\u6570\u636e\u3002Matplotlib.pyplot \u662f\u4e00\u4e2a\u7528\u4e8e\u521b\u5efa\u9759\u6001\u3001\u52a8\u753b\u548c\u4ea4\u4e92\u5f0f\u6570\u636e\u53ef\u89c6\u5316\u7684\u6d41\u884c\u5e93\uff0c\u5b83\u5efa\u7acb\u5728 Matplotlib \u7684\u57fa\u7840\u4e0a\u3002Seaborn \u53ef\u8ba9\u60a8\u4f7f\u7528 Matplotlib.pyplot \u8f7b\u677e\u521b\u5efa\u6781\u5177\u5438\u5f15\u529b\u7684\u7edf\u8ba1\u53ef\u89c6\u5316\u6548\u679c\u3002\u8fd9\u4e9b\u5e93\u4e3a\u5904\u7406\u6570\u636e\u3001\u6267\u884c\u5206\u6790\u548c\u751f\u6210\u53ef\u89c6\u5316\u6d1e\u5bdf\u529b\u63d0\u4f9b\u4e86\u5f3a\u5927\u7684\u57fa\u7840\u3002Path \u53ef\u5904\u7406\u6570\u636e\u6587\u4ef6\uff0cPandas \u548c Numpy \u53ef\u5904\u7406\u6570\u636e\u64cd\u4f5c\uff0c\u800c Seaborn \u548c Matplotlib \u53ef\u5c06\u7ed3\u679c\u53ef\u89c6\u5316\u3002<\/p>\n\n\n\n
from pyfolio.timeseries import perf_stats, extract_interesting_date_ranges\n# from pyfolio.tears import create_returns_tear_sheet<\/code><\/code><\/pre>\n\n\n\n
\u4e3a\u8fdb\u884c\u91d1\u878d\u6295\u8d44\u7ec4\u5408\u7ee9\u6548\u5206\u6790\uff0c\u8be5\u4ee3\u7801\u5bfc\u5165\u4e86 Pyfolio \u51fd\u6570\u3002\u4e3a\u8fdb\u884c\u4ea4\u6613\u7b56\u7565\u6027\u80fd\u5206\u6790\uff0cpyfolio.utils \u4e2d\u7684 extract_rets_pos_txn_from_zipline \u4ece Zipline \u56de\u6eaf\u6d4b\u8bd5\u7ed3\u679c\u4e2d\u63d0\u53d6\u6536\u76ca\u3001\u5934\u5bf8\u548c\u4ea4\u6613\u3002<\/p>\n\n\n\n

\u7136\u540e\u4ece pyfolio \u4e2d\u5bfc\u5165\u51e0\u4e2a\u7ed8\u56fe\u51fd\u6570\u3002plot_rolling_beta \u7528\u4e8e\u8861\u91cf\u6295\u8d44\u7ec4\u5408\u76f8\u5bf9\u4e8e\u57fa\u51c6\u7684\u6ce2\u52a8\u7387\uff0cplot_rolling_fama_french \u7528\u4e8e\u76f4\u89c2\u663e\u793a\u4e0e\u6cd5\u9a6c-\u6cd5\u5f0f\u56e0\u5b50\u76f8\u5173\u7684\u8868\u73b0\u3001plot_rolling_returns \u7528\u4e8e\u663e\u793a\u6eda\u52a8\u56de\u62a5\uff1bplot_rolling_sharpe \u7528\u4e8e\u8bc4\u4f30\u98ce\u9669\u8c03\u6574\u540e\u7684\u56de\u62a5\uff1bplot_rolling_volatility \u7528\u4e8e\u8bf4\u660e\u6295\u8d44\u7ec4\u5408\u6ce2\u52a8\u7387\u7684\u53d8\u5316\uff1bplot_drawdown_periods \u7528\u4e8e\u7a81\u51fa\u663e\u793a\u7f29\u6c34\u671f\uff1bplot_drawdown_underwater \u7528\u4e8e\u76f4\u89c2\u663e\u793a\u7f29\u6c34\u671f\u95f4\u6295\u8d44\u7ec4\u5408\u5728\u6c34\u4e0b\u7684\u7a0b\u5ea6\u3002<\/p>\n\n\n\n
\u4ece pyfolio.timeseries \u4e2d\uff0cperf_stats \u53ef\u4ee5\u8ba1\u7b97\u5173\u952e\u7684\u6027\u80fd\u6307\u6807\uff0c\u800c extract_interesting_date_ranges \u5219\u53ef\u4ee5\u8bc6\u522b\u7279\u5b9a\u7684\u65e5\u671f\uff0c\u4ee5\u4fbf\u6df1\u5165\u5206\u6790\u6027\u80fd\u7684\u6781\u7aef\u503c\u6216\u7f3a\u70b9\u3002\u540c\u6837\uff0ccreate_returns_tear_sheet \u4e5f\u88ab\u6ce8\u91ca\u6389\u4e86\uff0c\u8fd9\u8868\u660e\u5b83\u5728\u4e3a\u4e00\u7ec4\u56de\u62a5\u751f\u6210\u8be6\u7ec6\u7684\u6027\u80fd\u6307\u6807\u62a5\u544a\u65b9\u9762\u53ef\u80fd\u5f88\u6709\u7528\uff0c\u5c3d\u7ba1\u76ee\u524d\u5e76\u4e0d\u9700\u8981\u5b83\u3002<\/p>\n\n\n\n
%matplotlib inline\nplt.style.use('fivethirtyeight')\nwarnings.filterwarnings('ignore')<\/code><\/code><\/pre>\n\n\n\n
\u8be5\u4ee3\u7801\u6bb5\u4f7f\u7528 Matplotlib \u914d\u7f6e Jupyter Notebook \u73af\u5883\u4ee5\u5b9e\u73b0\u6570\u636e\u53ef\u89c6\u5316\u3002\u4f7f\u7528 %matplotlib inline \u547d\u4ee4\u53ef\u4f7f\u56fe\u8868\u76f4\u63a5\u663e\u793a\u5728\u7b14\u8bb0\u672c\u4e2d\uff0c\u65b9\u4fbf\u67e5\u770b\u53ef\u89c6\u5316\u8f93\u51fa\u3002\u547d\u4ee4 plt.style.use(‘fivethirtyeight’) \u4e3a\u56fe\u8868\u5e94\u7528\u4e86\u7c7b\u4f3c\u4e8e FiveThirtyEight \u7f51\u7ad9\u7684\u6837\u5f0f\uff0c\u4ee5\u7b80\u6d01\u73b0\u4ee3\u7684\u5916\u89c2\u589e\u5f3a\u4e86\u89c6\u89c9\u5438\u5f15\u529b\u3002\u6b64\u5916\uff0cwarnings.filterwarnings(‘ignore’) \u4f1a\u6291\u5236\u53ef\u80fd\u5e72\u6270\u8f93\u51fa\u7684\u8b66\u544a\u4fe1\u606f\uff0c\u4ece\u800c\u5728\u63a2\u7d22\u6027\u6570\u636e\u5206\u6790\u6216\u539f\u578b\u8bbe\u8ba1\u65f6\u63d0\u4f9b\u66f4\u5e72\u51c0\u7684\u5de5\u4f5c\u7a7a\u95f4\u3002\u8fd9\u4e9b\u547d\u4ee4\u4e3a\u53ef\u89c6\u5316\u8fc7\u7a0b\u63d0\u4f9b\u4e86\u4fbf\u5229\uff0c\u5e76\u589e\u5f3a\u4e86\u6570\u636e\u5206\u6790\u7684\u7f8e\u611f\u3002<\/p>\n\n\n\n
with pd.HDFStore('..\/01_trading_zipline\/backtests.h5') as store:\n    backtest = store['backtest']\nbacktest.info()<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u4f7f\u7528 pandas \u5e93\u8bbf\u95ee HDF5 \u6587\u4ef6\u4e2d\u7684\u6570\u636e\uff0c\u7279\u522b\u662f\u540d\u4e3a backtest \u7684\u6570\u636e\u96c6\u3002\u5b83\u4f7f\u7528 with \u8bed\u53e5\u9ad8\u6548\u5730\u6253\u5f00\u548c\u5173\u95ed HDF5 \u6587\u4ef6\uff0c\u5e76\u4f7f\u7528 pd.HDFStore \u51fd\u6570\u7ba1\u7406\u4f4d\u4e8e .\/.01_trading_zipline\/backtests.h5 \u7684\u6587\u4ef6\u3002\u5728\u8fd9\u79cd\u60c5\u51b5\u4e0b\uff0c\u56de\u6eaf\u6d4b\u8bd5\u6570\u636e\u96c6\u88ab\u52a0\u8f7d\u5230 backtest \u53d8\u91cf\u4e2d\u3002<\/p>\n\n\n\n
backtest.info()\u65b9\u6cd5\u7684\u8f93\u51fa\u603b\u7ed3\u4e86 DataFrame \u7684\u7ed3\u6784\uff0c\u663e\u793a\u4e86\u4ece 2010 \u5e74 1 \u6708 4 \u65e5\u5230 2017 \u5e74 12 \u6708 29 \u65e5\u7684 DatetimeIndex\uff0c\u5171\u6709 2013 \u4e2a\u6761\u76ee\u3002\u5b83\u5305\u62ec 30 \u5217\uff0c\u5206\u522b\u4ee3\u8868\u4e0e\u56de\u6eaf\u6d4b\u8bd5\u7ed3\u679c\u76f8\u5173\u7684\u5404\u79cd\u6307\u6807\uff0c\u5982\u7b97\u6cd5\u5468\u671f\u56de\u62a5\u7387\u3001\u963f\u5c14\u6cd5\u3001\u57fa\u51c6\u5468\u671f\u56de\u62a5\u7387\u548c\u8d85\u989d\u56de\u62a5\u7387\uff0c\u8fd9\u4e9b\u6307\u6807\u5bf9\u8bc4\u4f30\u4ea4\u6613\u7b56\u7565\u6027\u80fd\u81f3\u5173\u91cd\u8981\u3002<\/p>\n\n\n\n
\u5217\u6570\u636e\u7c7b\u578b\u5404\u4e0d\u76f8\u540c\uff0c\u5927\u591a\u6570\u4e3a float64\uff0c\u8868\u793a\u6570\u503c\uff0c\u800c\u5176\u4ed6\u5219\u4e3a int64 \u6216\u5bf9\u8c61\uff0c\u53ef\u80fd\u5305\u542b\u5206\u7c7b\u6570\u636e\u6216\u65f6\u95f4\u6233\u3002\u5927\u591a\u6570\u5217\u4e2d\u5b58\u5728\u975e\u7a7a\u6761\u76ee\u8868\u660e\u6570\u636e\u96c6\u662f\u5b8c\u6574\u7684\uff0c\u8fd9\u5bf9\u51c6\u786e\u5206\u6790\u81f3\u5173\u91cd\u8981\u3002\u8f93\u51fa\u8fd8\u63d0\u4f9b\u4e86 DataFrame \u7684\u5185\u5b58\u4f7f\u7528\u60c5\u51b5\uff08629.1 KB\uff09\uff0c\u8fd9\u5bf9\u4e86\u89e3\u6570\u636e\u96c6\u7684\u590d\u6742\u6027\u548c\u8fdb\u4e00\u6b65\u5206\u6790\u6240\u9700\u7684\u8ba1\u7b97\u8d44\u6e90\u81f3\u5173\u91cd\u8981\u3002<\/p>\n\n\n\n
returns, positions, transactions = extract_rets_pos_txn_from_zipline(backtest)<\/code><\/code><\/pre>\n\n\n\n
\u8fd9\u884c\u4ee3\u7801\u8c03\u7528\u51fd\u6570 extract_rets_pos_txn_from_zipline\uff0c\u5e76\u5c06\u53d8\u91cf backtest \u4f20\u9012\u7ed9\u5b83\u3002\u51fd\u6570\u5c06\u8fd4\u56de\u4e09\u4e2a\u503c\uff1a\u6536\u76ca\u3001\u4ed3\u4f4d\u548c\u4ea4\u6613\u3002\u6536\u76ca\u5305\u542b\u4e0e\u56de\u6eaf\u6d4b\u8bd5\u7684\u4ea4\u6613\u7b56\u7565\u76f8\u5173\u7684\u6027\u80fd\u6570\u636e\uff0c\u4ed3\u4f4d\u63d0\u4f9b\u4e86\u56de\u6eaf\u6d4b\u8bd5\u671f\u95f4\u4e0d\u540c\u65f6\u95f4\u6bb5\u6240\u6301\u80a1\u7968\u6216\u5408\u7ea6\u7684\u4fe1\u606f\uff0c\u800c\u4ea4\u6613\u5219\u8be6\u7ec6\u8bf4\u660e\u4e86\u6240\u6267\u884c\u7684\u4ea4\u6613\uff0c\u5305\u62ec\u4e70\u5165\u548c\u5356\u51fa\u3002\u8be5\u529f\u80fd\u53ef\u4ece\u56de\u6eaf\u6d4b\u8bd5\u4e2d\u63d0\u53d6\u91cd\u8981\u6307\u6807\u548c\u6570\u636e\uff0c\u4ece\u800c\u5bf9\u7ed3\u679c\u8fdb\u884c\u6709\u6548\u8bc4\u4f30\u548c\u5206\u6790\u3002<\/p>\n\n\n\n
returns.head().append(returns.tail())<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5 returns.head().append(returns.tail()) \u63d0\u4f9b\u4e86\u4e00\u4e2a\u5305\u542b\u8d22\u52a1\u6536\u76ca\u6570\u636e\u7684 DataFrame \u7684\u7b80\u660e\u6982\u89c8\u3002head() \u65b9\u6cd5\u83b7\u53d6\u521d\u59cb\u884c\uff0c\u800c tail() \u65b9\u6cd5\u83b7\u53d6\u6700\u7ec8\u884c\u3002\u901a\u8fc7\u8ffd\u52a0\u8fd9\u4e24\u7ec4\u6570\u636e\uff0c\u4ee3\u7801\u7ed3\u5408\u4e86\u6700\u65e9\u548c\u6700\u8fd1\u7684\u6761\u76ee\uff0c\u63d0\u4f9b\u4e86\u56de\u62a5\u7387\u968f\u65f6\u95f4\u53d8\u5316\u7684\u5feb\u7167\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u7ed3\u679c\u663e\u793a\u4e86\u4ece 2010 \u5e74 1 \u6708 4 \u65e5\u5230 2017 \u5e74 12 \u6708 29 \u65e5\u6309\u65e5\u671f\u7d22\u5f15\u7684\u56de\u62a5\u503c\u3002\u6700\u521d\u51e0\u884c\u53cd\u6620\u7684\u662f 2010 \u5e74 1 \u6708\u521d\u7684\u56de\u62a5\u7387\uff0c\u6570\u503c\u4ece 0.000000 \u5230 -0.003291\u30022017 \u5e74\u5e95\u7684\u6700\u540e\u4e00\u4e2a\u6761\u76ee\u663e\u793a\u4e86\u6b63\u8d1f\u6536\u76ca\u7684\u6df7\u5408\uff0c\u6700\u7ec8\u503c\u4e3a-0.005241\u3002\u8be5\u8f93\u51fa\u8bf4\u660e\u4e86\u56de\u62a5\u7387\u7684\u6ce2\u52a8\u60c5\u51b5\uff0c\u63d0\u4f9b\u4e86\u5bf9\u89c2\u5bdf\u671f\u5f00\u59cb\u548c\u7ed3\u675f\u65f6\u8d44\u4ea7\u8868\u73b0\u7684\u6df1\u5165\u4e86\u89e3\u3002<\/p>\n\n\n\n
\u8fd9\u79cd\u65b9\u6cd5\u80fd\u6709\u6548\u5e2e\u52a9\u5206\u6790\u5e08\u5feb\u901f\u8bc4\u4f30\u6295\u8d44\u4e1a\u7ee9\u8d8b\u52bf\uff0c\u663e\u793a\u4ece\u6570\u636e\u96c6\u5f00\u59cb\u5230\u7ed3\u675f\u7684\u56de\u62a5\u53d8\u5316\u60c5\u51b5\u3002<\/p>\n\n\n\n
positions.columns = [c.symbol for c in positions.columns[:-1]] + ['cash']\npositions.index = positions.index.normalize()\npositions.info()<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u4fee\u6539\u4e86\u4e00\u4e2a\u540d\u4e3a positions \u7684 DataFrame\uff0c\u5176\u4e2d\u5305\u542b\u4e00\u4e2a\u4ea4\u6613\u7b56\u7565\u7684\u8d22\u52a1\u6570\u636e\u3002\u9996\u5148\u66f4\u65b0\u5217\u540d\uff0c\u5c06\u6700\u540e\u4e00\u5217\u66ff\u6362\u4e3a “\u73b0\u91d1”\uff0c\u8868\u793a\u5b83\u4ee3\u8868\u73b0\u91d1\u6301\u6709\u91cf\uff0c\u8fd9\u662f\u7ba1\u7406\u8d22\u52a1\u6570\u636e\u7684\u5e38\u89c1\u505a\u6cd5\u3002\u7136\u540e\uff0c\u5c06 DataFrame \u7684\u7d22\u5f15\u5f52\u4e00\u5316\uff0c\u4ece\u4efb\u4f55\u65e5\u671f\u65f6\u95f4\u6761\u76ee\u4e2d\u79fb\u9664\u65f6\u95f4\u6210\u5206\uff0c\u4ee5\u4fbf\u4e8e\u65e5\u671f\u6bd4\u8f83\u3002<\/p>\n\n\n\n
\u4ee3\u7801\u8f93\u51fa\u4e86\u4fee\u6539\u540e\u7684 DataFrame \u7684\u6458\u8981\uff0c\u8be5 DataFrame \u7c7b\u578b\u4e3a pandas.core.frame.DataFrame\uff0c\u5305\u542b\u4ece 2010 \u5e74 1 \u6708 5 \u65e5\u5230 2017 \u5e74 12 \u6708 29 \u65e5\u7684 2,012 \u4e2a\u6761\u76ee\u3002\u5176\u4e2d\u5305\u62ec 1717 \u5217\uff0c\u4e3b\u8981\u662f\u8d44\u4ea7\u7b26\u53f7\u548c\u65b0\u7684 “\u73b0\u91d1 “\u5217\u3002\u6240\u6709\u6761\u76ee\u90fd\u662f float64 \u6570\u636e\u7c7b\u578b\uff0c\u8fd9\u8868\u660e\u5b83\u4eec\u4ee3\u8868\u7684\u662f\u8d27\u5e01\u6570\u503c\u3002DataFrame \u7684\u5185\u5b58\u4f7f\u7528\u91cf\u4e3a 26.4 MB\uff0c\u8868\u660e\u5176\u8d44\u6e90\u6d88\u8017\u60c5\u51b5\u3002\u8be5\u4ee3\u7801\u901a\u8fc7\u786e\u4fdd\u5217\u540d\u548c\u7d22\u5f15\u7684\u6b63\u786e\u683c\u5f0f\u5316\uff0c\u6709\u6548\u5730\u4e3a DataFrame \u7684\u5206\u6790\u6216\u53ef\u89c6\u5316\u505a\u597d\u4e86\u51c6\u5907\u3002<\/p>\n\n\n\n
transactions.symbol = transactions.symbol.apply(lambda x: x.symbol)<\/code><\/code><\/pre>\n\n\n\n
\u8fd9\u884c\u4ee3\u7801\u5728\u4e8b\u52a1 DataFrame \u7684\u7b26\u53f7\u5217\u4e0a\u4f7f\u7528 apply \u51fd\u6570\uff0c\u901a\u8fc7\u5e94\u7528 lambda \u51fd\u6570\u8f6c\u6362\u6bcf\u4e2a\u6761\u76ee\u3002lambda \u51fd\u6570\u63a5\u6536\u4e00\u4e2a\u53c2\u6570\uff0c\u5e76\u4ece\u7b26\u53f7\u5217\u4e2d\u7684\u6bcf\u4e2a\u5bf9\u8c61\u4e2d\u63d0\u53d6\u7b26\u53f7\u5c5e\u6027\u3002\u56e0\u6b64\uff0c\u6267\u884c\u8fd9\u4e00\u884c\u540e\uff0c\u7b26\u53f7\u5217\u5c06\u53ea\u5305\u542b\u6bcf\u4e2a\u539f\u59cb\u6761\u76ee\u7684\u7b26\u53f7\u5c5e\u6027\u503c\u3002\u5f53\u539f\u59cb\u6761\u76ee\u662f\u5bf9\u8c61\u65f6\uff0c\u8fd9\u79cd\u8f6c\u6362\u975e\u5e38\u6709\u7528\uff0c\u53ef\u4ee5\u5c06\u5176\u7b26\u53f7\u503c\u4fdd\u7559\u4e3a\u5b57\u7b26\u4e32\u6216\u5176\u4ed6\u7c7b\u578b\u3002<\/p>\n\n\n\n
transactions.head().append(transactions.tail())<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5 transactions.head().append(transactions.tail()) \u5c06\u540d\u4e3a transactions \u7684 DataFrame \u7684\u524d\u4e94\u884c\u548c\u540e\u4e94\u884c\u5408\u5e76\u5728\u4e00\u8d77\u3002\u901a\u8fc7\u8fd9\u4e00\u64cd\u4f5c\u53ef\u4ee5\u5feb\u901f\u68c0\u67e5\u6570\u636e\u96c6\u7684\u5f00\u5934\u548c\u7ed3\u5c3e\uff0c\u6709\u52a9\u4e8e\u8bc6\u522b\u8d8b\u52bf\u6216\u5f02\u5e38\u3002\u751f\u6210\u7684 DataFrame \u5305\u62ec\u91d1\u989d\u3001\u4f63\u91d1\u3001\u65e5\u671f\u3001\u8ba2\u5355\u53f7\u3001\u4ef7\u683c\u3001SID\u3001\u7b26\u53f7\u548c txn_dollars \u7b49\u5217\uff0c\u6570\u636e\u65f6\u95f4\u4e3a 2010 \u5e74 1 \u6708 5 \u65e5\u81f3 2017 \u5e74 12 \u6708 29 \u65e5\u3002\u6bcf\u4e00\u884c\u4ee3\u8868\u4e00\u7b14\u4ea4\u6613\uff0c\u8be6\u7ec6\u8bb0\u5f55\u4e86\u4ea4\u6613\u91d1\u989d\u3001\u4f63\u91d1\u3001\u65e5\u671f\u548c\u65f6\u95f4\u3001\u8ba2\u5355 ID\u3001\u4ef7\u683c\u548c\u8bc1\u5238\u4ee3\u7801\u3002<\/p>\n\n\n\n
\u4f8b\u5982\uff0c2010 \u5e74 1 \u6708 5 \u65e5\u7684\u7b2c\u4e00\u7b14\u4ea4\u6613\u663e\u793a\u91d1\u989d\u4e3a 1342\uff0c\u4ef7\u683c\u4e3a 37.630001\uff0c\u800c 2017 \u5e74 12 \u6708 29 \u65e5\u7684\u6700\u540e\u4e00\u7b14\u4ea4\u6613\u663e\u793a\u91d1\u989d\u4e3a -562\uff0c\u4ef7\u683c\u4e3a 23.090000\u3002\u8fd9\u79cd\u5934\u5c3e\u7ed3\u5408\u7684\u65b9\u5f0f\u53ef\u4ee5\u8ba9\u7528\u6237\u540c\u65f6\u67e5\u770b\u521d\u59cb\u4ea4\u6613\u548c\u6700\u540e\u4ea4\u6613\uff0c\u63d0\u4f9b\u6570\u636e\u8303\u56f4\u7684\u5feb\u7167\uff0c\u5e76\u6df1\u5165\u4e86\u89e3\u4ea4\u6613\u968f\u65f6\u95f4\u7684\u6f14\u53d8\u3002\u8f93\u51fa\u8bf4\u660e\u4e86\u4ea4\u6613 DataFrame \u7684\u7ed3\u6784\u548c\u5185\u5bb9\uff0c\u4fbf\u4e8e\u5206\u6790\u6574\u4e2a\u65f6\u95f4\u8f74\u4e0a\u7684\u8d22\u52a1\u6570\u636e\u3002<\/p>\n\n\n\n
HDF_PATH = Path('..', '..', 'data', 'assets.h5')<\/code><\/code><\/pre>\n\n\n\n
\u8fd9\u884c\u4ee3\u7801\u5b9a\u4e49\u4e86\u4e00\u4e2a\u540d\u4e3a HDF_PATH \u7684\u53d8\u91cf\uff0c\u7528\u4e8e\u5b58\u50a8\u540d\u4e3a assets.h5 \u7684\u6587\u4ef6\u8def\u5f84\u3002Path \u51fd\u6570\u6765\u81ea pathlib \u6a21\u5757\uff0c\u53ef\u4ee5\u8f7b\u677e\u64cd\u4f5c\u6587\u4ef6\u8def\u5f84\u3002\u901a\u8fc7\u4f7f\u7528 Path(‘…’,’…’,’data’,’assets.h5′)\uff0c\u5b83\u521b\u5efa\u4e86\u4e00\u4e2a\u76f8\u5bf9\u8def\u5f84\uff0c\u5411\u4e0a\u5bfc\u822a\u4e24\u4e2a\u76ee\u5f55\uff0c\u8fdb\u5165\u5305\u542b assets.h5 \u6587\u4ef6\u7684\u6570\u636e\u6587\u4ef6\u5939\u3002\u8fd9\u79cd\u65b9\u6cd5\u907f\u514d\u4e86\u786c\u7f16\u7801\u7684\u5b8c\u6574\u8def\u5f84\uff0c\u4ece\u800c\u589e\u5f3a\u4e86\u4ee3\u7801\u7684\u53ef\u79fb\u690d\u6027\uff0c\u56e0\u4e3a\u5728\u4e0d\u540c\u7684\u73af\u5883\u4e0b\uff0c\u5b8c\u6574\u8def\u5f84\u53ef\u80fd\u4f1a\u6709\u6240\u4e0d\u540c\u3002\u56e0\u6b64\uff0cHDF_PATH \u53ef\u4ee5\u7075\u6d3b\u3001\u7b80\u6d01\u5730\u6307\u5b9a assets.h5 \u6587\u4ef6\u7684\u4f4d\u7f6e\u3002<\/p>\n\n\n\n
assets = positions.columns[:-1]\nwith pd.HDFStore(HDF_PATH) as store:\n    df = store.get('us_equities\/stocks')['sector'].dropna()\n    df = df[~df.index.duplicated()]\nsector_map = df.reindex(assets).fillna('Unknown').to_dict()<\/code><\/code><\/pre>\n\n\n\n
\u8be5\u4ee3\u7801\u7ba1\u7406\u5b58\u50a8\u5728 HDF5 \u6587\u4ef6\u4e2d\u7684\u4e00\u7ec4\u8d44\u4ea7\u7684\u6247\u533a\u6570\u636e\u3002\u9996\u5148\uff0c\u5b83\u8981\u8bc6\u522b\u4f4d\u7f6e\u6570\u636e\u5e27\uff08DataFrame\uff09\u4e2d\u7684\u8d44\u4ea7\u5217\uff0c\u4f46\u4e0d\u5305\u62ec\u6700\u540e\u4e00\u5217\u3002\u7136\u540e\uff0c\u4f7f\u7528\u4e0a\u4e0b\u6587\u7ba1\u7406\u5668\u6253\u5f00\u4f4d\u4e8e HDF_PATH \u7684 HDF5 \u5b58\u50a8\uff0c\u4ee5\u786e\u4fdd\u4f7f\u7528\u540e\u6b63\u786e\u5173\u95ed\u3002\u5728\u6b64\u4e0a\u4e0b\u6587\u4e2d\uff0c\u5b83\u4f1a\u4ece us_equities\/stocks \u5173\u952e\u5b57\u4e2d\u68c0\u7d22 “\u6247\u533a “DataFrame\uff0c\u5e76\u5220\u9664\u6247\u533a\u5217\u4e2d\u4efb\u4f55\u7f3a\u5931\u503c\u7684\u6761\u76ee\uff0c\u4ee5\u6e05\u7406\u6570\u636e\u3002<\/p>\n\n\n\n
\u7136\u540e\uff0c\u901a\u8fc7\u8fc7\u6ee4\u552f\u4e00\u7d22\u5f15\u8bb0\u5f55\uff0c\u6d88\u9664\u6570\u636e\u5e27\u4e2d\u7684\u91cd\u590d\u7d22\u5f15\u3002\u5bf9\u6570\u636e\u5e27\u8fdb\u884c\u91cd\u65b0\u7d22\u5f15\uff0c\u4f7f\u5176\u4e0e\u8d44\u4ea7\u5217\u8868\u76f8\u5339\u914d\uff0c\u4efb\u4f55\u7f3a\u4e4f\u76f8\u5e94\u6247\u533a\u6761\u76ee\u7684\u8d44\u4ea7\u90fd\u4f1a\u88ab\u8d4b\u503c\u4e3a NaN\uff0c\u7136\u540e\u66ff\u6362\u4e3a “\u672a\u77e5”\u3002\u6700\u540e\uff0c\u5904\u7406\u540e\u7684 DataFrame \u4f1a\u8f6c\u6362\u4e3a\u5b57\u5178\uff0c\u5c06\u6bcf\u9879\u8d44\u4ea7\u6620\u5c04\u5230\u76f8\u5e94\u7684\u90e8\u95e8\uff0c\u65b9\u4fbf\u5feb\u901f\u67e5\u627e\uff0c\u5e76\u5c06\u672a\u627e\u5230\u7684\u8d44\u4ea7\u9ed8\u8ba4\u4e3a “\u672a\u77e5”\u3002<\/p>\n\n\n\n
with pd.HDFStore(HDF_PATH) as store:\n    benchmark_rets = store['sp500\/prices'].close.pct_change()\nbenchmark_rets.name = 'S&P500'\nbenchmark_rets = benchmark_rets.tz_localize('UTC').filter(returns.index)\nbenchmark_rets.tail()<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u4f7f\u7528 pandas \u5e93\u5904\u7406\u4ee5 HDF5 \u683c\u5f0f\u5b58\u50a8\u7684\u91d1\u878d\u6570\u636e\u3002\u5b83\u8bbf\u95ee\u4f4d\u4e8e HDF_PATH \u7684 HDF5 \u5b58\u50a8\uff0c\u4ee5\u68c0\u7d22\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u7684\u6536\u76d8\u4ef7\uff0c\u5e76\u5e94\u7528 pct_change \u65b9\u6cd5\u8ba1\u7b97\u8fde\u7eed\u6536\u76d8\u4ef7\u4e4b\u95f4\u7684\u767e\u5206\u6bd4\u53d8\u5316\uff0c\u4ece\u800c\u5c06\u4ef7\u683c\u6570\u636e\u8f6c\u6362\u4e3a\u6536\u76ca\uff0c\u8fd9\u5bf9\u91d1\u878d\u5206\u6790\u81f3\u5173\u91cd\u8981\u3002<\/p>\n\n\n\n
\u8ba1\u7b97\u6536\u76ca\u540e\uff0c\u4ee3\u7801\u4f1a\u5c06\u751f\u6210\u7684\u7cfb\u5217\u6807\u8bb0\u4e3a “S&P500″\uff0c\u4ee5\u4fbf\u4e8e\u8bc6\u522b\u3002\u5b83\u4f7f\u7528 tz_localize(‘UTC’)\u65b9\u6cd5\u5c06\u65f6\u95f4\u6233\u6807\u51c6\u5316\u4e3a UTC\u3002filter(returns.index)\u65b9\u6cd5\u5c06\u57fa\u51c6\u6536\u76ca\u7387\u4e0e\u53e6\u4e00\u4e2a\u540d\u4e3a returns \u7684\u7cfb\u5217\u5bf9\u9f50\uff0c\u786e\u4fdd benchmark_rets \u7cfb\u5217\u4e2d\u53ea\u4fdd\u7559\u6536\u76ca\u7387\u7d22\u5f15\u4e2d\u7684\u65e5\u671f\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u663e\u793a\u4e86 benchmark_rets \u7cfb\u5217\u7684\u6700\u540e\u51e0\u4e2a\u6761\u76ee\uff0c\u5176\u4e2d\u6355\u6349\u4e86\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u7684\u65e5\u671f\u548c\u76f8\u5e94\u7684\u6bcf\u65e5\u767e\u5206\u6bd4\u56de\u62a5\u7387\uff0c\u663e\u793a\u4e86\u6307\u5b9a\u65f6\u671f\u5185\u7684\u6b63\u8d1f\u6ce2\u52a8\u3002\u8be5\u8f93\u51fa\u8fd8\u786e\u8ba4\u56de\u62a5\u503c\u4e3a float64 \u7c7b\u578b\uff0c\u9002\u5408\u91d1\u878d\u5206\u6790\u4e2d\u7684\u6570\u503c\u8ba1\u7b97\u3002\u5728\u5c06\u9879\u76ee\u4ece Zipline \u8fc7\u6e21\u5230 Pyfolio \u7684\u8fc7\u7a0b\u4e2d\uff0c\u8fd9\u4e00\u8fc7\u7a0b\u5bf9\u4e8e\u98ce\u9669\u8bc4\u4f30\u6216\u4e1a\u7ee9\u6bd4\u8f83\u7b49\u8fdb\u4e00\u6b65\u5206\u6790\u81f3\u5173\u91cd\u8981\u3002<\/p>\n\n\n\n
perf_stats(returns=returns, factor_returns=benchmark_rets, positions=positions, transactions=transactions)<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u51fd\u6570 perf_stats \u901a\u8fc7\u5206\u6790\u63d0\u4f9b\u7684\u6536\u76ca\u3001\u57fa\u51c6\u6536\u76ca\u3001\u5934\u5bf8\u548c\u4ea4\u6613\u4e2d\u7684\u5404\u79cd\u8d22\u52a1\u6307\u6807\u6765\u8bc4\u4f30\u4ea4\u6613\u7b56\u7565\u7684\u6027\u80fd\u3002\u8f93\u51fa\u7ed3\u679c\u5305\u62ec\u8bc4\u4f30\u7b56\u7565\u6709\u6548\u6027\u548c\u98ce\u9669\u72b6\u51b5\u7684\u5173\u952e\u7ee9\u6548\u6307\u6807\u3002<\/p>\n\n\n\n
\u5e74\u6536\u76ca\u7387\u7ea6\u4e3a 0.0499\uff0c\u8868\u793a\u5e74\u5e73\u5747\u5229\u6da6\uff0c\u7d2f\u8ba1\u6536\u76ca\u7387\u7ea6\u4e3a 0.4754\uff0c\u53cd\u6620\u5206\u6790\u671f\u5185\u7684\u603b\u6536\u76ca\u3002\u5e74\u6ce2\u52a8\u7387\u7ea6\u4e3a 0.0859\uff0c\u7528\u4e8e\u8861\u91cf\u6536\u76ca\u53d8\u5316\u548c\u76f8\u5173\u98ce\u9669\u3002\u590f\u666e\u6bd4\u7387\u7ea6\u4e3a 0.6094\uff0c\u8868\u660e\u6bcf\u5355\u4f4d\u98ce\u9669\u7684\u5408\u7406\u56de\u62a5\u3002<\/p>\n\n\n\n
\u5361\u5c14\u9a6c\u6bd4\u7387\u7ea6\u4e3a 0.4557\uff0c\u5c06\u5e74\u5ea6\u56de\u62a5\u4e0e-0.1095 \u7684\u6700\u5927\u7f29\u6c34\u8fdb\u884c\u6bd4\u8f83\uff0c\u663e\u793a\u51fa\u6295\u8d44\u4ef7\u503c\u7684\u6700\u5927\u8dcc\u5e45\uff0c\u51f8\u663e\u51fa\u5de8\u5927\u7684\u6f5c\u5728\u635f\u5931\u3002\u7a33\u5b9a\u5ea6\u4e3a 0.9323\uff0c\u8868\u660e\u8868\u73b0\u7a33\u5b9a\u3002<\/p>\n\n\n\n
\u5176\u4ed6\u6307\u6807\u5305\u62ec 1.1227 \u7684 Omega \u6bd4\u7387\uff08\u8868\u660e\u6536\u76ca\u9ad8\u4e8e\u635f\u5931\u7684\u53ef\u80fd\u6027\uff09\u548c\u7ea6 0.8790 \u7684 Sortino \u6bd4\u7387\uff08\u5173\u6ce8\u4e0b\u884c\u98ce\u9669\uff09\u3002\u504f\u5ea6\u4e3a-0.0320\uff0c\u8868\u660e\u7565\u5fae\u503e\u5411\u4e8e\u8d1f\u6536\u76ca\uff0c\u800c\u5cf0\u5ea6\u4e3a 7.1644\uff0c\u8868\u660e\u51fa\u73b0\u6781\u7aef\u7ed3\u679c\u7684\u53ef\u80fd\u6027\u8f83\u9ad8\u3002<\/p>\n\n\n\n
\u7ea6 1.0144 \u7684\u5c3e\u90e8\u6bd4\u7387\u8bc4\u4f30\u4e86\u6781\u7aef\u635f\u5931\u7684\u98ce\u9669\uff0c-0.0166 \u7684\u6bcf\u65e5\u98ce\u9669\u503c\u91cf\u5316\u4e86\u6f5c\u5728\u7684\u6bcf\u65e5\u635f\u5931\u3002\u603b\u6760\u6746\u7387\u4e3a 1.4014\uff0c\u53cd\u6620\u4e86\u501f\u5165\u8d44\u91d1\u7684\u7a0b\u5ea6\uff0c\u65e5\u5468\u8f6c\u7387\u7ea6\u4e3a 0.0404\uff0c\u8868\u660e\u4e86\u4ea4\u6613\u9891\u7387\u3002\u963f\u5c14\u6cd5\u503c\u4e3a-0.0072\uff0c\u8868\u660e\u8868\u73b0\u4e0d\u4f73\uff0c\u800c\u8d1d\u5854\u503c\u4e3a 0.4948\uff0c\u8868\u660e\u4e0e\u5e02\u573a\u8d70\u52bf\u6709\u4e00\u5b9a\u7684\u76f8\u5173\u6027\u3002\u8fd9\u4e9b\u6307\u6807\u5408\u5728\u4e00\u8d77\uff0c\u53ef\u4ee5\u5168\u9762\u53cd\u6620\u4ea4\u6613\u7b56\u7565\u7684\u7ee9\u6548\u548c\u98ce\u9669\u7279\u5f81\u3002<\/p>\n\n\n\n
plot_perf_stats(returns=returns, factor_returns=benchmark_rets)<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
plot_perf_stats(returns=returns, factor_returns=benchmark_rets)\u4ee3\u7801\u7247\u6bb5\u53ef\u89c6\u5316\u8d22\u52a1\u7b56\u7565\u6216\u6295\u8d44\u7ec4\u5408\u7684\u7ee9\u6548\u7edf\u8ba1\u6570\u636e\u3002\u5b83\u6709\u4e24\u4e2a\u53c2\u6570\uff1areturns \u4ee3\u8868\u6240\u5206\u6790\u6295\u8d44\u7684\u6536\u76ca\uff0cfactor_returns \u4ee3\u8868\u7528\u4e8e\u6bd4\u8f83\u7684\u57fa\u51c6\u6216\u5e02\u573a\u56e0\u5b50\u7684\u6536\u76ca\u3002<\/p>\n\n\n\n
\u8be5\u529f\u80fd\u53ef\u751f\u6210\u4e00\u4e2a\u65b9\u6846\u56fe\uff0c\u663e\u793a\u5404\u79cd\u7ee9\u6548\u6307\u6807\uff0c\u5982\u5e74\u56de\u62a5\u7387\u3001\u7d2f\u8ba1\u56de\u62a5\u7387\u3001\u5e74\u6ce2\u52a8\u7387\u3001\u590f\u666e\u6bd4\u7387\u3001\u5361\u5c14\u9a6c\u6bd4\u7387\u3001\u6700\u5927\u7f29\u6c34\u7387\u548c\u963f\u5c14\u6cd5\u7b49\u3002\u901a\u8fc7\u8fd9\u4e9b\u6307\u6807\u53ef\u4ee5\u6df1\u5165\u4e86\u89e3\u6295\u8d44\u4e1a\u7ee9\u7684\u4e0d\u540c\u65b9\u9762\u3002\u4f8b\u5982\uff0c\u5e74\u5ea6\u56de\u62a5\u7387\u663e\u793a\u7684\u662f\u5e73\u5747\u5e74\u5ea6\u56de\u62a5\u7387\uff0c\u800c\u590f\u666e\u6bd4\u7387\u8bc4\u4f30\u7684\u662f\u98ce\u9669\u8c03\u6574\u540e\u7684\u56de\u62a5\u7387\u3002<\/p>\n\n\n\n
\u7bb1\u5f62\u56fe\u7684\u5f62\u5f0f\u53ef\u4ee5\u6e05\u695a\u5730\u6bd4\u8f83\u6295\u8d44\u4e0e\u5176\u57fa\u51c6\u4e4b\u95f4\u7684\u8fd9\u4e9b\u6307\u6807\u3002\u65b9\u6846\u8868\u793a\u56db\u5206\u4f4d\u6570\u4e4b\u95f4\u7684\u8303\u56f4\uff0c\u6bcf\u4e2a\u65b9\u6846\u5185\u7684\u7ebf\u8868\u793a\u4e2d\u503c\uff0c\u800c\u987b\u7ebf\u8868\u793a\u6570\u636e\u8303\u56f4\uff0c\u5e76\u76f8\u5e94\u5730\u6807\u51fa\u5f02\u5e38\u503c\u3002\u8fd9\u79cd\u53ef\u89c6\u5316\u65b9\u5f0f\u6709\u52a9\u4e8e\u8bc4\u4f30\u6295\u8d44\u76f8\u5bf9\u4e8e\u57fa\u51c6\u7684\u98ce\u9669\u548c\u6536\u76ca\u60c5\u51b5\uff0c\u4fbf\u4e8e\u8bc6\u522b\u4e1a\u7ee9\u7684\u4f18\u52bf\u548c\u52a3\u52bf\u3002\u603b\u4f53\u800c\u8a00\uff0c\u8be5\u529f\u80fd\u5bf9\u6295\u8d44\u8005\u548c\u5206\u6790\u5e08\u8bc4\u4f30\u548c\u6bd4\u8f83\u6295\u8d44\u7b56\u7565\u975e\u5e38\u6709\u7528\u3002<\/p>\n\n\n\n
oos_date = '2017-01-01'<\/code><\/code><\/pre>\n\n\n\n
\u8fd9\u4e00\u884c\u521b\u5efa\u4e86\u4e00\u4e2a\u540d\u4e3a oos_date \u7684\u53d8\u91cf\uff0c\u5e76\u4e3a\u5176\u5206\u914d\u4e86\u5b57\u7b26\u4e32\u503c 2017-01-01\uff0c\u8868\u793a 2017 \u5e74 1 \u6708 1 \u65e5\u3002\u5982\u679c\u4ee5\u540e\u9700\u8981\u8fdb\u884c\u65e5\u671f\u64cd\u4f5c\uff0c\u53ef\u80fd\u9700\u8981\u4f7f\u7528 Python \u4e2d\u7684 datetime \u7b49\u5e93\u5c06\u6b64\u5b57\u7b26\u4e32\u8f6c\u6362\u4e3a\u65e5\u671f\u5bf9\u8c61\u3002\u73b0\u5728\uff0c\u5b83\u53ea\u662f\u4e00\u4e2a\u7b80\u5355\u7684\u65e5\u671f\u5b57\u7b26\u4e32\u3002<\/p>\n\n\n\n
show_perf_stats(returns=returns, \n                factor_returns=benchmark_rets, \n                positions=positions, \n                transactions=transactions, \n                live_start_date=oos_date)<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u8c03\u7528 show_perf_stats \u51fd\u6570\u6765\u5206\u6790\u548c\u663e\u793a\u4ea4\u6613\u7b56\u7565\u7684\u6027\u80fd\u7edf\u8ba1\u3002\u5b83\u9700\u8981\u7684\u53c2\u6570\u5305\u62ec\uff1a\u7531\u7b56\u7565\u4ea7\u751f\u7684\u6536\u76ca\uff1b\u7528\u4e8e\u57fa\u51c6\u6bd4\u8f83\u7684 factor_returns\uff1b\u4ea4\u6613\u671f\u95f4\u6301\u6709\u7684\u4ed3\u4f4d\uff1b\u8bb0\u5f55\u5df2\u6267\u884c\u4ea4\u6613\u7684\u4e8b\u52a1\uff1b\u4ee5\u53ca\u6807\u8bb0\u6837\u672c\u5916\u6d4b\u8bd5\u5f00\u59cb\u7684 live_start_date\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u603b\u7ed3\u4e86\u6837\u672c\u5185\u548c\u6837\u672c\u5916\u4e24\u4e2a\u9636\u6bb5\u7684\u7ee9\u6548\u6307\u6807\uff0c\u6837\u672c\u5185\u8de8\u5ea6\u4e3a 83 \u4e2a\u6708\uff0c\u6837\u672c\u5916\u8de8\u5ea6\u4e3a 11 \u4e2a\u6708\u3002\u5e74\u5ea6\u56de\u62a5\u7387\u7565\u6709\u63d0\u9ad8\uff0c\u4ece\u6837\u672c\u5185\u7684 5.0% \u63d0\u9ad8\u5230\u6837\u672c\u5916\u7684 5.6%\uff0c\u8868\u660e\u540e\u4e00\u9636\u6bb5\u7684\u8868\u73b0\u66f4\u597d\u3002\u7d2f\u8ba1\u56de\u62a5\u7387\u4e5f\u5448\u73b0\u7c7b\u4f3c\u8d8b\u52bf\uff0c\u6837\u672c\u5185\u56de\u62a5\u7387\u4e3a 47.5%\uff0c\u6837\u672c\u5916\u56de\u62a5\u7387\u4e3a 39.8%\u3002<\/p>\n\n\n\n
\u5e74\u6ce2\u52a8\u7387\u4ece\u6837\u672c\u5185\u7684 8.6% \u964d\u81f3\u6837\u672c\u5916\u7684 4.9%\uff0c\u8868\u660e\u672a\u89c1\u6570\u636e\u7684\u6ce2\u52a8\u7387\u964d\u4f4e\u3002\u590f\u666e\u6bd4\u7387\u4ece 0.61 \u63d0\u9ad8\u5230 1.13\uff0c\u8868\u660e\u6837\u672c\u5916\u671f\u95f4\u7684\u98ce\u9669\u8c03\u6574\u56de\u62a5\u7387\u66f4\u9ad8\u3002\u5361\u5c14\u9a6c\u6bd4\u7387\u4e5f\u4ece 0.46 \u663e\u8457\u4e0a\u5347\u5230 2.09\uff0c\u53cd\u6620\u4e86\u7ee9\u6548\u7684\u63d0\u9ad8\u3002<\/p>\n\n\n\n
\u6700\u5927\u7f29\u6c34\u7387\u7b49\u6307\u6807\u663e\u793a\uff0c\u6837\u672c\u5916\u4e1a\u7ee9\u4ece\u6837\u672c\u5185\u7684-10.9%\u4e0b\u964d\u5230-2.7%\uff0c\u8868\u660e\u540e\u4e00\u65f6\u671f\u7684\u5f39\u6027\u66f4\u5927\u3002\u5176\u4ed6\u6307\u6807\uff0c\u5305\u62ec Omega \u6bd4\u7387\u548c Sortino \u6bd4\u7387\uff0c\u5728\u6837\u672c\u5916\u6570\u636e\u4e2d\u663e\u793a\u51fa\u6709\u5229\u7684\u98ce\u9669\u72b6\u51b5\u3002<\/p>\n\n\n\n
\u603b\u4e4b\uff0c\u7814\u7a76\u7ed3\u679c\u8868\u660e\uff0c\u5728\u6839\u636e\u6837\u672c\u5916\u6570\u636e\u8fdb\u884c\u8bc4\u4f30\u65f6\uff0c\u4ea4\u6613\u7b56\u7565\u4e0d\u4ec5\u4fdd\u6301\u4e86\u5176\u6027\u80fd\uff0c\u800c\u4e14\u5728\u51e0\u4e2a\u5173\u952e\u9886\u57df\u6709\u6240\u6539\u8fdb\uff0c\u8fd9\u51f8\u663e\u4e86\u5176\u7a33\u5065\u6027\u548c\u5728\u73b0\u5b9e\u4e16\u754c\u4e2d\u7684\u5e94\u7528\u6f5c\u529b\u3002<\/p>\n\n\n\n
plot_rolling_returns(returns=returns, factor_returns=benchmark_rets, live_start_date=oos_date, cone_std=(1.0, 1.5, 2.0))\nplt.gcf().set_size_inches(14, 8)<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u4f7f\u7528 plot_rolling_returns \u51fd\u6570\u53ef\u89c6\u5316\u4ea4\u6613\u7b56\u7565\u76f8\u5bf9\u4e8e\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u7684\u8868\u73b0\u3002\u5b83\u9700\u8981\u7b56\u7565\u6536\u76ca\u3001\u57fa\u51c6\u6536\u76ca\u548c\u5b9e\u65f6\u4ea4\u6613\u5f00\u59cb\u65e5\u671f\u7b49\u53c2\u6570\u3002con_std \u53c2\u6570\u8bbe\u7f6e\u4e3a 1.0\u30011.5 \u548c 2.0\uff0c\u7528\u4e8e\u5b9a\u4e49\u56fe\u4e2d\u7f6e\u4fe1\u533a\u95f4\u7684\u6807\u51c6\u504f\u5dee\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u663e\u793a\u4e86\u4e00\u6bb5\u65f6\u95f4\u5185\u7684\u7d2f\u8ba1\u56de\u62a5\uff0c\u4e09\u6761\u7ebf\u4ee3\u8868\u4e0d\u540c\u7684\u7ee9\u6548\u6307\u6807\u3002\u7070\u8272\u7ebf\u8868\u793a 2010 \u5e74\u81f3 2018 \u5e74\u7684\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u57fa\u51c6\u3002\u7eff\u7ebf\u8868\u793a\u4ea4\u6613\u7b56\u7565\u7684\u56de\u6eaf\u6d4b\u8bd5\u6536\u76ca\u7387\uff0c\u8be5\u7b56\u7565\u6700\u521d\u843d\u540e\u4e8e\u57fa\u51c6\uff0c\u4f46\u5728\u968f\u540e\u51e0\u5e74\u4e2d\u8868\u73b0\u51fa\u663e\u8457\u589e\u957f\u3002\u7ea2\u7ebf\u53cd\u6620\u7684\u662f\u5b9e\u65f6\u56de\u62a5\uff0c\u5b83\u7d27\u8ddf\u56de\u6eaf\u6d4b\u8bd5\u7684\u56de\u62a5\uff0c\u4f46\u4e5f\u6709\u4e00\u4e9b\u6ce2\u52a8\u3002<\/p>\n\n\n\n
\u5b9e\u65f6\u6536\u76ca\u7ebf\u5468\u56f4\u7684\u9634\u5f71\u533a\u57df\u8bf4\u660e\u4e86\u7f6e\u4fe1\u533a\u95f4\uff0c\u663e\u793a\u4e86\u57fa\u4e8e\u5386\u53f2\u6ce2\u52a8\u7684\u9884\u671f\u6536\u76ca\u8303\u56f4\u3002\u8fd9\u79cd\u53ef\u89c6\u5316\u65b9\u5f0f\u6709\u52a9\u4e8e\u5c06\u7b56\u7565\u7684\u8868\u73b0\u4e0e\u57fa\u51c6\u8fdb\u884c\u6bd4\u8f83\uff0c\u5f3a\u8c03\u4ea4\u6613\u7b56\u7565\u5728\u4e00\u6bb5\u65f6\u95f4\u5185\u7684\u6709\u6548\u6027\u300214 x 8 \u82f1\u5bf8\u7684\u7248\u9762\u5c3a\u5bf8\u786e\u4fdd\u4e86\u53ef\u8bfb\u6027\uff0c\u4f7f\u5176\u6210\u4e3a\u76f8\u5bf9\u4e8e\u5e02\u573a\u8fdb\u884c\u7ee9\u6548\u5206\u6790\u7684\u91cd\u8981\u5de5\u5177\u3002<\/p>\n\n\n\n
plot_rolling_sharpe(returns=returns)\nplt.gcf().set_size_inches(14, 8);<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5 plot_rolling_sharpe(returns=returns) \u663e\u793a\u4e00\u7ec4\u6536\u76ca\u968f\u7740\u65f6\u95f4\u63a8\u79fb\u7684\u6eda\u52a8\u590f\u666e\u6bd4\u7387\u3002\u590f\u666e\u6bd4\u7387\u8861\u91cf\u98ce\u9669\u8c03\u6574\u540e\u7684\u6536\u76ca\uff0c\u663e\u793a\u76f8\u5bf9\u4e8e\u98ce\u9669\u8f83\u9ad8\u8d44\u4ea7\u7684\u6ce2\u52a8\u7387\u7684\u8d85\u989d\u6536\u76ca\u3002\u8be5\u51fd\u6570\u8ba1\u7b97\u7684\u662f\u516d\u4e2a\u6708\u6eda\u52a8\u7a97\u53e3\u5185\u7684\u590f\u666e\u6bd4\u7387\uff0c\u6709\u52a9\u4e8e\u6df1\u5165\u4e86\u89e3\u4e0d\u540c\u5e02\u573a\u6761\u4ef6\u4e0b\u7684\u8868\u73b0\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u56fe\u50cf\u663e\u793a\u4e86 2011 \u5e74\u81f3 2018 \u5e74\u7684\u6eda\u52a8\u590f\u666e\u6bd4\u7387\u3002\u6a59\u8272\u7ebf\u63cf\u7ed8\u4e86\u8ba1\u7b97\u51fa\u7684\u590f\u666e\u6bd4\u7387\uff0c\u5728\u8d85\u8fc7 4 \u7684\u5730\u65b9\u6709\u660e\u663e\u7684\u5cf0\u503c\uff0c\u8868\u660e\u98ce\u9669\u8c03\u6574\u540e\u7684\u56de\u62a5\u7387\u5f88\u9ad8\u3002\u84dd\u8272\u865a\u7ebf\u8868\u793a\u4f5c\u4e3a\u57fa\u51c6\u7684\u5e73\u5747\u590f\u666e\u6bd4\u7387\uff0c\u800c\u4f4d\u4e8e\u96f6\u70b9\u7684\u9ed1\u7ebf\u5219\u8868\u793a\u76f8\u5bf9\u4e8e\u98ce\u9669\u7684\u5145\u8db3\u56de\u62a5\u7684\u4e34\u754c\u503c\u3002\u56fe\u4e2d\u663e\u793a\u4e86\u590f\u666e\u6bd4\u7387\u4f4e\u4e8e\u96f6\u7684\u60c5\u51b5\uff0c\u8868\u660e\u5728\u6b64\u671f\u95f4\u6536\u76ca\u4e0d\u8db3\u4ee5\u8bc1\u660e\u98ce\u9669\u7684\u5408\u7406\u6027\u3002\u8fd9\u79cd\u53ef\u89c6\u5316\u65b9\u5f0f\u6709\u6548\u5730\u8bf4\u660e\u4e86\u6295\u8d44\u6218\u7565\u7684\u7ee9\u6548\uff0c\u63ed\u793a\u4e86\u6210\u529f\u4e0e\u6311\u6218\u3002<\/p>\n\n\n\n
plot_rolling_beta(returns=returns, factor_returns=benchmark_rets)\nplt.gcf().set_size_inches(14, 8);<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u51fd\u6570 plot_rolling_beta \u663e\u793a\u6295\u8d44\u7ec4\u5408\u4e0e\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u57fa\u51c6\u76f8\u6bd4\u7684\u6eda\u52a8\u8d1d\u5854\u7cfb\u6570\u3002\u53c2\u6570 returns \u548c factor_returns \u5206\u522b\u4ee3\u8868\u6295\u8d44\u7ec4\u5408\u548c\u57fa\u51c6\u6536\u76ca\u3002\u6eda\u52a8\u8d1d\u5854\u503c\u8861\u91cf\u7684\u662f\u6295\u8d44\u7ec4\u5408\u6536\u76ca\u968f\u57fa\u51c6\u6536\u76ca\u53d8\u5316\u800c\u53d8\u5316\u7684\u7a0b\u5ea6\u3002<\/p>\n\n\n\n
\u8f93\u51fa\u7ed3\u679c\u4e3a\u65f6\u95f4\u5e8f\u5217\u56fe\uff0cX \u8f74\u4e3a 2011 \u5e74\u81f3 2018 \u5e74\u7684\u5e74\u4efd\uff0cY \u8f74\u4e3a\u8d1d\u5854\u503c\u3002\u56fe\u4e2d\u663e\u793a\u4e86\u4e24\u6761\u7ebf\uff1a\u84dd\u8272\u5b9e\u7ebf\u8868\u793a 6 \u4e2a\u6708\u7684\u6eda\u52a8\u8d1d\u5854\u503c\uff0c\u6d45\u7070\u8272\u7ebf\u8868\u793a 12 \u4e2a\u6708\u7684\u6eda\u52a8\u8d1d\u5854\u503c\u3002\u6c34\u5e73\u865a\u7ebf 0.5 \u8868\u793a\u8d1d\u5854\u503c\uff0c\u8868\u660e\u6295\u8d44\u7ec4\u5408\u7684\u6ce2\u52a8\u6027\u4f4e\u4e8e\u57fa\u51c6\u3002<\/p>\n\n\n\n
\u8d1d\u5854\u503c\u5728\u6574\u4e2a\u7ed8\u5236\u671f\u95f4\u6ce2\u52a8\uff0c\u53cd\u6620\u4e86\u6295\u8d44\u7ec4\u5408\u76f8\u5bf9\u4e8e\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u7684\u98ce\u9669\u72b6\u51b5\u7684\u53d8\u5316\u30026 \u4e2a\u6708\u7684\u8d1d\u5854\u503c\u66f4\u80fd\u53cd\u6620\u8fd1\u671f\u7684\u5e02\u573a\u72b6\u51b5\uff0c\u800c 12 \u4e2a\u6708\u7684\u8d1d\u5854\u503c\u5219\u63d0\u4f9b\u4e86\u66f4\u5e73\u6ed1\u7684\u957f\u671f\u89c6\u89d2\u3002\u8fd9\u79cd\u53ef\u89c6\u5316\u65b9\u6cd5\u6709\u52a9\u4e8e\u6295\u8d44\u8005\u4e86\u89e3\u5176\u6295\u8d44\u7ec4\u5408\u7684\u98ce\u9669\u655e\u53e3\u968f\u65f6\u95f4\u7684\u53d8\u5316\uff0c\u5c24\u5176\u662f\u5728\u4e0d\u540c\u7684\u5e02\u573a\u9636\u6bb5\uff0c\u5e76\u652f\u6301\u6839\u636e\u5386\u53f2\u8868\u73b0\u505a\u51fa\u660e\u667a\u7684\u6295\u8d44\u51b3\u7b56\u3002<\/p>\n\n\n\n
fig, ax = plt.subplots(nrows=2, ncols=2, figsize=(16, 10))\naxes = ax.flatten()<\/code><\/code><\/pre>\n\n\n\n
plot_drawdown_periods(returns=returns, ax=axes[0])\nplot_rolling_beta(returns=returns, factor_returns=benchmark_rets, ax=axes[1])\nplot_drawdown_underwater(returns=returns, ax=axes[2])\nplot_rolling_sharpe(returns=returns)\nplt.tight_layout();<\/code><\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u4f7f\u7528 Matplotlib \u521b\u5efa\u4e86\u4e00\u4e2a 2×2 \u7f51\u683c\u7684\u5b50\u56fe\uff0c\u4ee5\u53ef\u89c6\u5316\u91d1\u878d\u6295\u8d44\u7ec4\u5408\u968f\u65f6\u95f4\u53d8\u5316\u7684\u7ee9\u6548\u6307\u6807\u3002\u7b2c\u4e00\u4e2a\u5b50\u56fe\u663e\u793a\u4e86\u524d 10 \u4e2a\u7f29\u6c34\u671f\uff0c\u84dd\u7ebf\u4ee3\u8868 2010 \u5e74\u81f3 2018 \u5e74\u7684\u7d2f\u8ba1\u56de\u62a5\uff0c\u9634\u5f71\u533a\u57df\u8868\u793a\u5927\u5e45\u7f29\u6c34\u3002\u8fd9\u6709\u52a9\u4e8e\u8bc6\u522b\u5927\u5e45\u4e0b\u8dcc\u671f\u548c\u8bc4\u4f30\u98ce\u9669\u655e\u53e3\u3002<\/p>\n\n\n\n
\u7b2c\u4e8c\u4e2a\u5b50\u56fe\u663e\u793a\u6295\u8d44\u7ec4\u5408\u4e0e\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u5728 6 \u4e2a\u6708\u548c 12 \u4e2a\u6708\u671f\u95f4\u7684\u6eda\u52a8\u8d1d\u5854\u503c\u3002\u8d1d\u5854\u503c\u663e\u793a\u4e86\u6295\u8d44\u7ec4\u5408\u5bf9\u5e02\u573a\u6ce2\u52a8\u7684\u654f\u611f\u5ea6\uff0c\u8d1d\u5854\u503c\u8d85\u8fc7 1 \u8868\u793a\u6ce2\u52a8\u7387\u9ad8\u4e8e\u5e02\u573a\u3002\u8be5\u56fe\u8bf4\u660e\u4e86\u6295\u8d44\u7ec4\u5408\u7684\u98ce\u9669\u72b6\u51b5\u662f\u5982\u4f55\u968f\u65f6\u95f4\u53d8\u5316\u7684\u3002<\/p>\n\n\n\n
\u7b2c\u4e09\u4e2a\u5b50\u56fe\u662f\u6c34\u4e0b\u56fe\uff0c\u7a81\u51fa\u663e\u793a\u4e86\u6307\u5b9a\u65f6\u95f4\u6bb5\u5185\u4ece\u9ad8\u5cf0\u5230\u4f4e\u8c37\u7684\u6700\u5927\u635f\u5931\u3002\u6a59\u8272\u7ebf\u8868\u793a\u7f29\u6c34\u6df1\u5ea6\uff0c\u8ba9\u4eba\u4e86\u89e3\u6295\u8d44\u7ec4\u5408\u8868\u73b0\u6700\u5dee\u7684\u65f6\u671f\u3002<\/p>\n\n\n\n
\u7b2c\u56db\u4e2a\u5b50\u56fe\u663e\u793a\u7684\u662f\u6309 6 \u4e2a\u6708\u95f4\u9694\u8ba1\u7b97\u7684\u6eda\u52a8\u590f\u666e\u6bd4\u7387\uff0c\u8be5\u6bd4\u7387\u8861\u91cf\u6295\u8d44\u7ec4\u5408\u7684\u98ce\u9669\u8c03\u6574\u540e\u56de\u62a5\u3002\u590f\u666e\u6bd4\u7387\u8d8a\u9ad8\uff0c\u8868\u660e\u4e1a\u7ee9\u4e0e\u98ce\u9669\u76f8\u6bd4\u8d8a\u597d\uff0c\u8be5\u56fe\u53cd\u6620\u4e86\u590f\u666e\u6bd4\u7387\u968f\u65f6\u95f4\u7684\u6ce2\u52a8\u3002<\/p>\n\n\n\n
\u8fd9\u4e9b\u53ef\u89c6\u5316\u5185\u5bb9\u5171\u540c\u63d0\u4f9b\u4e86\u6295\u8d44\u7ec4\u5408\u7684\u4e1a\u7ee9\u3001\u98ce\u9669\u548c\u56de\u62a5\u7279\u5f81\u7684\u5168\u9762\u6982\u89c8\uff0c\u6709\u52a9\u4e8e\u505a\u51fa\u66f4\u660e\u667a\u7684\u6295\u8d44\u51b3\u7b56\u3002<\/p>\n\n\n\n
interesting_times = extract_interesting_date_ranges(returns=returns)
(interesting_times['Fall2015']
 .to_frame('pf').join(benchmark_rets)
 .add(1).cumprod().sub(1)
 .plot(lw=2, figsize=(14, 6), title='Post-Brexit Turmoil'))
plt.tight_layout()<\/code><\/pre>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u4ee3\u7801\u7247\u6bb5\u4ece\u6536\u76ca\u6570\u636e\u96c6\u4e2d\u63d0\u53d6\u7279\u5b9a\u65e5\u671f\u8303\u56f4\uff0c\u4e3b\u8981\u96c6\u4e2d\u5728 2016 \u5e74\u82f1\u56fd\u8131\u6b27\u6295\u7968\u540e\u7684\u65f6\u671f\u3002\u51fd\u6570 extract_interesting_date_ranges \u8c03\u7528\u7684\u662f\u56de\u62a5\u6570\u636e\uff08\u53ef\u80fd\u5305\u542b\u5386\u53f2\u8d22\u52a1\u4e1a\u7ee9\u6570\u636e\uff09\uff0c\u7ed3\u679c\u5b58\u50a8\u5728 interesting_times \u53d8\u91cf\u4e2d\uff0c\u4e3b\u8981\u96c6\u4e2d\u5728 2015 \u5e74\u79cb\u5b63\u3002<\/p>\n\n\n\n
\u7136\u540e\u5bf9\u63d0\u53d6\u7684\u6570\u636e\u8fdb\u884c\u8f6c\u6362\uff0c\u4ee5\u63d0\u9ad8\u53ef\u7528\u6027\u3002to_frame \u65b9\u6cd5\u5c06\u6240\u9009\u6570\u636e\u8f6c\u6362\u4e3a\u6807\u6709 “pf “\u7684 DataFrame\uff0c\u5e76\u5c06\u5176\u4e0e\u53e6\u4e00\u4e2a DataFrame\uff08benchmark_rets\uff0c\u5305\u542b\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u6536\u76ca\u7b49\u57fa\u51c6\u6536\u76ca\u6570\u636e\uff09\u8fde\u63a5\u8d77\u6765\u3002\u5408\u5e76\u540e\u7684\u6570\u636e\u4f1a\u8fdb\u884c\u8ba1\u7b97\uff0c\u5c06\u6bcf\u4e2a\u503c\u52a0 1\uff0c\u8ba1\u7b97\u7d2f\u8ba1\u4e58\u79ef\uff0c\u518d\u51cf\u53bb 1\uff0c\u4ece\u800c\u6709\u6548\u8ba1\u7b97\u51fa\u6307\u5b9a\u671f\u95f4\u7684\u7d2f\u8ba1\u56de\u62a5\uff0c\u7528\u4e8e\u4e1a\u7ee9\u6bd4\u8f83\u3002<\/p>\n\n\n\n
\u6700\u540e\u4e00\u6b65\u662f\u7ed8\u5236\u7ed3\u679c\u56fe\uff0c\u4ee5 2 \u7684\u7ebf\u5bbd\u548c\u6307\u5b9a\u7684\u6570\u5b57\u5927\u5c0f\u76f4\u89c2\u663e\u793a\u6295\u8d44\u7ec4\u5408\u5728\u4e00\u6bb5\u65f6\u95f4\u5185\u76f8\u5bf9\u4e8e\u57fa\u51c6\u7684\u7d2f\u8ba1\u56de\u62a5\u3002\u6807\u9898\u4e3a “\u82f1\u56fd\u8131\u6b27\u540e\u7684\u52a8\u8361”\uff0c\u53cd\u6620\u4e86\u8fd9\u4e00\u91cd\u5927\u653f\u6cbb\u4e8b\u4ef6\u4e2d\u7684\u5e02\u573a\u884c\u4e3a\u3002\u8f93\u51fa\u56fe\u50cf\u663e\u793a\u4e86\u4e24\u6761\u7ebf\uff1a\u4e00\u6761\u662f\u84dd\u8272\u7684\u6295\u8d44\u7ec4\u5408\u8868\u73b0\u7ebf\uff0c\u53e6\u4e00\u6761\u662f\u7ea2\u8272\u7684\u6807\u51c6\u666e\u5c14 500 \u6307\u6570\u7ebf\uff0c\u8bf4\u660e\u4e86\u82f1\u56fd\u8131\u6b27\u540e\u671f\u95f4\u56de\u62a5\u7684\u6ce2\u52a8\u6027\u548c\u8d8b\u52bf\uff0c\u5e76\u7a81\u51fa\u4e86\u6295\u8d44\u7ec4\u5408\u76f8\u5bf9\u4e8e\u5927\u76d8\u7684\u8868\u73b0\u3002<\/p>\n\n\n\n
\u89c2\u70b9\u56de\u987e<\/strong><\/h3>\n\n\n\n
    \n
  1. \u7b97\u6cd5\u4ea4\u6613\u7b56\u7565\u7684\u5f00\u53d1\u548c\u6d4b\u8bd5\u5e94\u8be5\u57fa\u4e8e\u5386\u53f2\u6570\u636e\u548c\u73b0\u5b9e\u7684\u4ea4\u6613\u6761\u4ef6<\/strong>\uff0c\u4ee5\u786e\u4fdd\u7b56\u7565\u5728\u5b9e\u9645\u5e02\u573a\u4e2d\u7684\u6709\u6548\u6027\u3002<\/li>\n\n\n\n
  2. \u5747\u503c\u56de\u5f52\u662f\u7b97\u6cd5\u4ea4\u6613\u4e2d\u7684\u4e00\u79cd\u91cd\u8981\u7b56\u7565<\/strong>\uff0c\u5b83\u5047\u8bbe\u8d44\u4ea7\u4ef7\u683c\u4f1a\u56de\u5f52\u5230\u5176\u5e73\u5747\u6c34\u5e73\uff0c\u56e0\u6b64\u53ef\u4ee5\u901a\u8fc7\u8bc6\u522b\u504f\u79bb\u5e73\u5747\u6c34\u5e73\u7684\u8d44\u4ea7\u6765\u83b7\u5229\u3002<\/li>\n\n\n\n
  3. \u7ee9\u6548\u5206\u6790\u548c\u53ef\u89c6\u5316\u5bf9\u4e8e\u8bc4\u4f30\u4ea4\u6613\u7b56\u7565\u81f3\u5173\u91cd\u8981<\/strong>\uff0cPyfolio \u63d0\u4f9b\u4e86\u4e00\u5957\u5b8c\u6574\u7684\u5de5\u5177\u6765\u5206\u6790\u548c\u5c55\u793a\u7b56\u7565\u7684\u7ee9\u6548\u6307\u6807\u3002<\/li>\n\n\n\n
  4. \u6570\u636e\u7684\u51c6\u786e\u5904\u7406\u548c\u5b58\u50a8\u5bf9\u4e8e\u7b97\u6cd5\u4ea4\u6613\u81f3\u5173\u91cd\u8981<\/strong>\uff0c\u4f7f\u7528 HDF5 \u6587\u4ef6\u53ef\u4ee5\u6709\u6548\u5730\u7ba1\u7406\u548c\u8bbf\u95ee\u5927\u578b\u6570\u636e\u96c6\u3002<\/li>\n\n\n\n
  5. \u98ce\u9669\u7ba1\u7406\u662f\u7b97\u6cd5\u4ea4\u6613\u4e2d\u7684\u4e00\u4e2a\u5173\u952e\u73af\u8282<\/strong>\uff0c\u901a\u8fc7\u5206\u6790\u6ce2\u52a8\u7387\u3001\u590f\u666e\u6bd4\u7387\u3001\u6700\u5927\u56de\u64a4\u548c\u8d1d\u5854\u503c\u7b49\u6307\u6807\uff0c\u53ef\u4ee5\u66f4\u597d\u5730\u7406\u89e3\u548c\u63a7\u5236\u4ea4\u6613\u98ce\u9669\u3002<\/li>\n\n\n\n
  6. \u4f7f\u7528\u6eda\u52a8\u7a97\u53e3\u6765\u5206\u6790\u548c\u76d1\u63a7\u7b56\u7565\u7684\u77ed\u671f\u548c\u957f\u671f\u98ce\u9669\u655e\u53e3<\/strong>\uff0c\u5982\u6eda\u52a8\u8d1d\u5854\u548c\u6eda\u52a8\u590f\u666e\u6bd4\u7387\uff0c\u6709\u52a9\u4e8e\u65f6\u523b\u638c\u63e1\u7b56\u7565\u7684\u98ce\u9669\u72b6\u51b5\u3002<\/li>\n\n\n\n
  7. \u7b97\u6cd5\u4ea4\u6613\u7b56\u7565\u7684\u7ee9\u6548\u4e0d\u4ec5\u8981\u5728\u6837\u672c\u5185\u6570\u636e\u4e0a\u8fdb\u884c\u6d4b\u8bd5<\/strong>\uff0c\u8fd8\u5e94\u8be5\u5728\u6837\u672c\u5916\u6570\u636e\u4e0a\u8fdb\u884c\u9a8c\u8bc1\uff0c\u4ee5\u786e\u4fdd\u7b56\u7565\u7684\u7a33\u5065\u6027\u548c\u9002\u7528\u6027\u3002<\/li>\n\n\n\n
  8. \u901a\u8fc7\u53ef\u89c6\u5316\u5de5\u5177\uff0c\u5982 Matplotlib \u548c Seaborn<\/strong>\uff0c\u53ef\u4ee5\u66f4\u76f4\u89c2\u5730\u5c55\u793a\u7b56\u7565\u7684\u8868\u73b0\u548c\u5e02\u573a\u52a8\u6001\uff0c\u5e2e\u52a9\u4ea4\u6613\u8005\u505a\u51fa\u66f4\u660e\u667a\u7684\u51b3\u7b56\u3002<\/li>\n<\/ol>\n\n\n\n
    \u611f\u8c22\u60a8\u9605\u8bfb\u5230\u6700\u540e\u3002\u5982\u679c\u5bf9\u6587\u4e2d\u7684\u5185\u5bb9\u6709\u4efb\u4f55\u7591\u95ee\uff0c\u8bf7\u7ed9\u6211\u7559\u8a00\uff0c\u5fc5\u590d\u3002<\/strong><\/p>\n\n\n\n
    \n\n\n\n
    <\/p>\n\n\n\n
    <\/p>\n\n\n\n
    \u672c<\/mark><\/strong>\u6587\u5185\u5bb9\u4ec5\u4ec5\u662f\u6280\u672f\u63a2\u8ba8\u548c\u5b66\u4e60\uff0c\u5e76\u4e0d\u6784\u6210\u4efb\u4f55\u6295\u8d44\u5efa\u8bae\u3002<\/mark><\/strong><\/p>\n\n\n\n
    \u8f6c\u53d1\u8bf7\u6ce8\u660e\u539f\u4f5c\u8005\u548c\u51fa\u5904\u3002<\/mark><\/strong><\/p>\n","protected":false},"excerpt":{"rendered":"
    \u4f5c\u8005\uff1a\u8001\u4f59\u635e\u9c7c \u539f\u521b\u4e0d\u6613\uff0c\u8f6c\u8f7d\u8bf7\u6807\u660e\u51fa\u5904\u53ca\u539f\u4f5c\u8005\u3002…<\/p>\n
    Continue reading\u5e72\u8d27\uff01Zipline \u4e0e Pyfolio \u643a\u624b\u73a9\u8f6c\u7b97\u6cd5\u4ea4\u6613\u548c\u56de\u6eaf\u6d4b\u8bd5<\/span><\/a><\/div>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[3,5],"class_list":{"0":"post-1588","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"hentry","6":"category-aiinvest","7":"tag-ai","9":"entry"},"_links":{"self":[{"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/posts\/1588","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/comments?post=1588"}],"version-history":[{"count":1,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/posts\/1588\/revisions"}],"predecessor-version":[{"id":1589,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/posts\/1588\/revisions\/1589"}],"wp:attachment":[{"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/media?parent=1588"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/categories?post=1588"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/laoyulaoyu.com\/index.php\/wp-json\/wp\/v2\/tags?post=1588"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}