{"id":27106,"date":"2021-06-21T21:47:28","date_gmt":"2021-06-21T13:47:28","guid":{"rendered":"https:\/\/swarma.org\/?p=27106"},"modified":"2021-06-21T21:47:28","modified_gmt":"2021-06-21T13:47:28","slug":"%e5%b9%82%e5%be%8b%e5%88%86%e5%b8%83%e6%8b%9f%e5%90%88%e7%a5%9e%e5%99%a8-python%e5%ba%93powerlaw","status":"publish","type":"post","link":"https:\/\/swarma.org\/?p=27106","title":{"rendered":"\u5e42\u5f8b\u5206\u5e03\u62df\u5408\u795e\u5668\u2014\u2014Python\u5e93powerlaw"},"content":{"rendered":"
\n
\n
\n
\n

<\/p>\n<\/section>\n

\n
\n
\n
\n
\n
\n

<\/section>\n<\/section>\n
\n
\n

\u5bfc\u8bed<\/strong><\/p>\n<\/section>\n<\/section>\n

\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
\n
\n

<\/strong>\u5177\u6709\u957f\u5c3e\u7279\u5f81\u7684\u5206\u5e03\u5f80\u5f80\u4e00\u76ee\u4e86\u7136\uff0c\u4f46\u5b9e\u9645\u62df\u5408\u8fc7\u7a0b\u5374\u53ef\u80fd\u9047\u5230\u5404\u79cd\u5404\u6837\u7684\u95ee\u9898\u3002\u672c\u6587\u5c06\u4e3a\u8bfb\u8005\u4ecb\u7ecd2014\u5e74\u7531\u65b0\u52a0\u5761\u79d1\u6280\u8bbe\u8ba1\u5927\u5b66\u548c\u9ebb\u7701\u7406\u5de5\u7814\u7a76\u8005\u8054\u5408\u53d1\u5e03\u7684python\u5e93\uff1apowerlaw\uff0c\u4e13\u95e8\u9002\u7528\u4e8e\u5e42\u5f8b\u7b49\u957f\u5c3e\u7279\u5f81\u5206\u5e03\u7684\u62df\u5408\uff0c\u89e3\u51b3\u62df\u5408\u70e6\u607c\u3002<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\n
\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n
<\/section>\n<\/section>\n
\n

\u80e1\u4e00\u51b0<\/span> | \u4f5c\u8005<\/strong><\/span><\/strong><\/p>\n<\/section>\n

\n

\u9093\u4e00\u96ea<\/span> | \u7f16\u8f91<\/strong><\/span><\/p>\n<\/section>\n


<\/p>\n


<\/section>\n

<\/p>\n

\n
\n
\n

\u671f\u520a\u6765\u6e90\uff1a<\/span>PLOS ONE<\/span><\/em><\/span><\/span><\/p>\n

\u8bba\u6587\u6807\u9898\uff1a<\/span><\/p>\n

powerlaw: A Python Package for Analysis of Heavy- Tailed Distributions<\/span>
<\/p>\n<\/section>\n<\/section>\n

\u8bba\u6587\u7f51\u5740\uff1a<\/span><\/p>\n

https:\/\/journals.plos.org\/plosone\/article?id=10.1371\/journal.pone.0085777<\/span><\/p>\n<\/blockquote>\n


<\/section>\n

<\/section>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

1. \u5e42\u5f8b\u5206\u5e03\u7b80\u8981\u56de\u987e<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n


<\/p>\n

\u5e42\u5f8b\u5206\u5e03\u4f5c\u4e3a\u957f\u5c3e\u5206\u5e03\u7684\u4e00\u79cd\uff0c\u6ee1\u8db3<\/span>\u53ef\u4ee5\u7528\u6765\u5370\u8bc1\u751f\u6d3b\u4e2d\u5f88\u591a\u6709\u8da3\u7684\u73b0\u8c61\uff0c\u6bd4\u5982\u6700\u4e3a\u5927\u5bb6\u719f\u77e5\u7684\u201c\u4e8c\u516b\u5b9a\u5f8b\u201d\u201c\u5bcc\u8005\u8d8a\u5bcc\u201d\u7b49\u7b49\u3002\u65e9\u5728\u4e0a\u4e16\u7eaa\u65e9\u671f\u5c31\u5148\u540e\u6709\u514b\u83b1\u4f2f\u5b9a\u5f8b\u3001Zipf\u5b9a\u5f8b\u7b49\u5e42\u5f8b\u73b0\u8c61\u7684\u53d1\u73b0\uff0c\u800c\u5343\u79a7\u5e74\u4e4b\u9645BA\u65e0\u6807\u5ea6\u7f51\u7edc\u7684\u6b63\u5f0f\u63d0\u51fa\u66f4\u662f\u6380\u8d77\u4e00\u6ce2\u201c\u5e42\u5f8b\u70ed\u6f6e\u201d\uff0c\u5927\u5bb6\u8d8a\u6765\u8d8a\u63a5\u53d7\u201c\u5e42\u5f8b\u5206\u5e03\u662f\u590d\u6742\u7cfb\u7edf\u4e2d\u7684\u666e\u9002\u73b0\u8c61\u201d\u8fd9\u4e00\u89c2\u70b9\u3002<\/span><\/p>\n

       <\/span>  <\/span><\/p>\n


<\/p>\n

\u56fe1. \u4eba\u4eec\u5e38\u8bf4\u201c20%\u7684\u4eba\u638c\u63e1\u774080%\u7684\u8d22\u5bcc\u201d\u5c31\u662f\u5178\u578b\u7684\u5e42\u5f8b\u5206\u5e03\uff0c\u5176\u5177\u6709\u5c16\u5cf0\u3001\u957f\u5c3e\u7279\u5f81\u3002<\/span><\/section>\n


<\/p>\n


<\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

2. powerlaw\u5e93\u62df\u5408\u6548\u679c<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n

       <\/span><\/p>\n

\u56fe2. \u4e09\u4e2a\u4e0d\u540c\u6570\u636e\u96c6\u7528powerlaw\u62df\u5408\u540e\u7684\u53ef\u89c6\u5316\u56fe\u50cf<\/span><\/p>\n


<\/p>\n

\u56fe2\u663e\u793a\u4e86\u53ef\u89c6\u5316\u3001\u62df\u5408\u548c\u8bc4\u4f30\u957f\u5c3e\u5206\u5e03\u7684\u57fa\u672c\u8981\u7d20\uff0c\u6bcf\u7ec4\u5b50\u56fe\u5c06\u5728\u540e\u7eed\u90e8\u5206\u4e2d\u8fdb\u884c\u8fdb\u4e00\u6b65\u7684\u8be6\u7ec6\u63cf\u8ff0\u3002<\/span><\/p>\n


<\/p>\n

\u56fe1\u4e2d\u4e09\u5217\u5b50\u56fe\u5305\u542b\u4e09\u4e2a\u793a\u4f8b\u6570\u636e\u96c6\u548c\u62df\u5408\u7ed3\u679c\uff0c\u5206\u522b\u8868\u793a\u5e42\u5f8b\u826f\u597d\u62df\u5408\u3001\u4e2d\u7b49\u62df\u5408\u548c\u8f83\u5dee\u62df\u5408\u3002\u7b2c\u4e00\u5217\u7684\u6570\u636e\uff0c\u4e5f\u662f\u6700\u597d\u7684\u62df\u5408\u6570\u636e\u96c6\u53ef\u80fd\u662f\u6240\u6709\u5e42\u5f8b\u5206\u5e03\u4e2d\u6700\u8457\u540d\u548c\u6700\u53ef\u9760\u7684\uff1a\u82f1\u8bed\u5355\u8bcd\u4f7f\u7528\u9891\u7387\u3002\u5177\u4f53\u4f7f\u7528\u7684\u6570\u636e\u662fHerman Melville\u7684\u5c0f\u8bf4\u300aMoby Dick\u300b\u4e2d\u8bcd\u8bed\u7684\u4f7f\u7528\u9891\u7387\u3002\u7b2c\u4e8c\u5217\u7684\u6570\u636e\u662f\u7ebf\u866b\u6bcf\u4e2a\u795e\u7ecf\u5143\u7684\u8fde\u63a5\u6570\u3002\u6700\u540e\u4e00\u4e2a\u62df\u5408\u4e0d\u4f73\u7684\u6570\u636e\u662f1984\u5e74\u81f32002\u5e74\u95f4\u53d7\u505c\u7535\u5f71\u54cd\u7684\u7f8e\u56fd\u4eba\u6570\u3002<\/span><\/p>\n


<\/p>\n

\u56fe2\u4e2d\u7684\u5b50\u56feA\u8868\u793a\u4e86\u4e09\u4e2a\u6837\u672c\u6570\u636e\u96c6\u7684\u6982\u7387\u5bc6\u5ea6\u51fd\u6570\uff0c\u5b50\u56feB\u663e\u793a\u4e86\u53ea\u6709\u5c11\u6570\u5206\u5e03\u7684\u5c3e\u90e8\u53ef\u80fd\u9075\u5faa\u5e42\u5f8b\uff0c\u5b50\u56feC\u8868\u793a\u4e86\u5982\u4f55\u5c06\u5e42\u5f8b\u62df\u5408\u4e0e\u5176\u4ed6\u957f\u5c3e\u5206\u5e03\u8fdb\u884c\u6bd4\u8f83\u3002<\/span><\/p>\n


<\/p>\n


<\/p>\n

\n
\n
\n
\n
\n

<\/section>\n

<\/section>\n
\n
<\/svg><\/section>\n<\/section>\n<\/section>\n
\n

2. powerlaw\u5e93\u57fa\u672c\u64cd\u4f5c\u4ecb\u7ecd<\/strong><\/strong><\/p>\n<\/section>\n

\n

<\/section>\n

<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/h3>\n


<\/p>\n

\n
\n
\n

\u53ef\u89c6\u5316<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n

\n


<\/p>\n<\/section>\n

\u4e0b\u9762\u4ee51984\u5e74\u81f32002\u5e74\u95f4\u53d7\u505c\u7535\u5f71\u54cd\u7684\u7f8e\u56fd\u4eba\u6570\u6570\u636e\u4e3a\u4f8b\u8fdb\u884c\u64cd\u4f5c\u8bf4\u660e\u3002<\/span><\/p>\n


<\/p>\n

\n
\n
><\/span>import<\/span> powerlaw
<\/span><\/section>\n
>fit = powerlaw.Fit(data)  <\/span>#\u5bf9\u5bfc\u5165\u7684\u6570\u636edata\u8fdb\u884c\u5e42\u5f8b\u62df\u5408<\/span>
>fit.power_law.alpha <\/span> #\u62df\u5408\u5f97\u5230\u5e42\u5f8b\u5206\u5e03\u5e42\u6307\u6570alpha<\/span><\/section>\n
2.273  <\/span>
>fit.distribution_compare(\u2018power_law\u2019, \u2018exponential\u2019)  <\/span>#\u4e0e\u6307\u6570\u5206\u5e03\u5bf9\u6bd4\u5f97\u5230\u5bf9\u6570\u4f3c\u7136\u6bd4\u548cp\u503c<\/span>
(<\/span>12.755, 0.152<\/span>)<\/span><\/section>\n<\/section>\n<\/section>\n


<\/p>\n

powerlaw\u5e93\u80fd\u5bb9\u6613\u7ed8\u5236\u6982\u7387\u5bc6\u5ea6\u51fd\u6570<\/span>\uff08PDF\uff09<\/span>\u3001\u7d2f\u79ef\u5206\u5e03\u51fd\u6570<\/span>\uff08CDF\uff09<\/span>\u548c\u4e92\u8865\u7d2f\u79ef\u5206\u5e03\u51fd\u6570<\/span>\uff08CCDF\uff09<\/span>\u3002\u5bf9\u5e94\u7684\u8ba1\u7b97\u5de5\u4f5c\u901a\u8fc7pdf\u3001cdf\u548cccdf\u5b8c\u6210\u7684\uff0c\u800c\u7ed8\u56fe\u5219\u662f\u7ed3\u5408matplotlib\u901a\u8fc7plot_pdf\u3001plot_cdf\u548cplot_ccdf\u547d\u4ee4\u5b8c\u6210\u3002<\/span><\/p>\n


<\/p>\n

\n
\n
>powerlaw.plot_pdf(data, color = \u2018b\u2019)  <\/span>
>powerlaw.plot_pdf(data, linear_bins = <\/span><\/span>True<\/span>, color = \u2018r\u2019)<\/span><\/section>\n<\/section>\n<\/section>\n


<\/p>\n

\u8be5\u547d\u4ee4\u4e2d\u7684linear_bins\u53c2\u6570\u63a7\u5236\u5bf9\u6570\u5750\u6807\u8f74\u4e0e\u5426\u3002\u7ebf\u6027\u523b\u5ea6\u4e0d\u5229\u4e8e\u5e42\u5f8b\u5c3e\u90e8\u201c\u5927\u503c\u201d\u7684\u89c2\u5bdf\uff0c\u91c7\u7528\u5bf9\u6570\u523b\u5ea6\u589e\u52a0\u5728\u5c3e\u90e8\u89c2\u5bdf\u4e00\u7cfb\u5217\u6570\u636e\u7684\u53ef\u80fd\u6027\u3002\u56fe2\u5b50\u56feA\u663e\u793a\u4e86\u9009\u62e9\u5bf9\u6570\u523b\u5ea6\u5982\u4f55\u6781\u5927\u5730\u63d0\u9ad8\u6570\u636e\u5206\u5e03\u7684\u53ef\u89c6\u5316\u3002\u5728\u8fd9\u91cc\u9ed8\u8ba4\u4e3a\u5bf9\u6570\u523b\u5ea6\uff0c\u4f46\u4e5f\u53ef\u4ee5\u7528linear_bins = True\u8fdb\u884c\u7ebf\u6027\u5212\u5206\u3002<\/span><\/p>\n


<\/p>\n

\u5173\u4e8ePDF\u3001CDF\u548cCCDF\u7684\u62df\u5408\u793a\u4f8b\u5982\u4e0b\uff0c\u5177\u4f53\u9009\u62e9\u9700\u7ed3\u5408\u62df\u5408\u9700\u6c42\u3002\u5982\u679c\u8981\u5728\u540c\u4e00\u753b\u5e03\u4e2d\u5448\u73b0\u591a\u6761\u62df\u5408\u66f2\u7ebf\uff0c\u9700\u8981\u4f20\u9012\u5e26\u6709ax\u7684matplotlib\u5bf9\u8c61\u3002\u4f8b\u5982\u4e0b\u65b9\u540c\u65f6\u753b\u51fa\u795e\u7ecf\u5143\u8fde\u63a5\u6570\u636e\u7684CCDF<\/span>\uff08\u7ea2\u8272\uff09<\/span>\u548cPDF<\/span>\uff08\u84dd\u8272\uff09<\/span>\u66f2\u7ebf\u3002<\/span><\/p>\n


<\/p>\n

\n
\n
>fig2 = fit.plot_pdf(color = \u2018b\u2019, linewidth = <\/span>2<\/span>)<\/span><\/section>\n
>fit.power_law.plot_pdf(color = \u2018b\u2019, linestyle = \u2018\u2013\u2019, ax = fig2)<\/span>
>fit.plot_ccdf(color = \u2018r\u2019, linewidth = <\/span><\/span>2<\/span>, ax = fig2)<\/span>
>fit.power_law.plot_ccdf(color = \u2018r\u2019, linestyle = \u2018\u2013\u2019, ax = fig2<\/span><\/span><\/section>\n<\/section>\n<\/section>\n


<\/p>\n

       <\/span>\u56fe3. \u795e\u7ecf\u5143\u8fde\u63a5\u6570\u636e\u7684CCDF\uff08\u7ea2\u8272\uff09\u548cPDF\uff08\u84dd\u8272\uff09\u66f2\u7ebf\u3002<\/span>
<\/p>\n


<\/p>\n

\u5728\u7ed8\u56fe\u4e4b\u5916\uff0cPDF\u3001CDF\u548cCCDF\u4fe1\u606f\u4e5f\u53ef\u7528\u3002\u62df\u5408\u5bf9\u8c61\u8fd4\u56de\u62df\u5408\u6570\u636e\u548c\u6392\u5e8f\u6570\u636e<\/span>\uff08CDF\uff09<\/span>\u548cbin\u7684\u8fb9\u7f18\u6982\u7387<\/span>\uff08PDF\uff09<\/span>\u3002\u5206\u5e03\u5bf9\u8c61\u9ed8\u8ba4\u4f7f\u7528\u5168\u90e8\u62df\u5408\u6570\u636e\uff0c\u4e5f\u53ef\u4ee5\u6307\u5b9a\u5177\u4f53\u7684\u6570\u636e\u8303\u56f4\u3002<\/span><\/p>\n


<\/p>\n

>x, y = fit.cdf() <\/span> #\u5e42\u5f8b\u5206\u5e03CDF\u62df\u5408<\/span>
>bin_edges, probability = fit.pdf()<\/span>
>y = fit.lognormal.cdf(data = [<\/span><\/span>300<\/span>, <\/span>350<\/span>])  <\/span>#\u5bf9\u6570\u6b63\u6001\u5206\u5e03CDF\u62df\u5408<\/span>
>y = fit.lognormal.pdf()<\/span><\/section><\/section><\/section><\/pre>\n

<\/span><\/strong><\/span><\/h4>\n
\n
\n
\n
\u62df\u5408\u8303\u56f4<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n

<\/p>\n<\/section>\n
\u62df\u5408\u5e42\u5f8b\u7684\u7b2c\u4e00\u6b65\u662f\u786e\u5b9a\u62df\u5408\u6570\u636e\u7684\u54ea\u4e00\u90e8\u5206\u3002\u957f\u5c3e\u5206\u5e03\u7684\u6709\u8da3\u7279\u5f81\u662f\u5c3e\u90e8\u53ca\u5176\u5c5e\u6027\uff0c\u6240\u4ee5\u5982\u679c\u6570\u636e\u7684\u521d\u59cb\u5c0f\u503c\u4e0d\u9075\u5faa\u5e42\u5f8b\u5206\u5e03\uff0c\u7528\u6237\u6709\u53ef\u80fd\u4f1a\u9009\u62e9\u5ffd\u7565\u5b83\u4eec\u3002powerlaw\u5e93\u4f1a\u4e3a\u7528\u6237\u63d0\u4f9b\u6700\u4f18\u7684\u5e42\u5f8b\u62df\u5408\u6700\u5c0f\u503c\uff0c\u5f53\u7136\u7528\u6237\u4e5f\u53ef\u4ee5\u9009\u62e9\u6307\u5b9a\u6700\u5c0f\u503c\u6216\u7ed9\u6700\u5c0f\u503c\u6846\u5b9a\u8303\u56f4\uff0c\u4e0d\u540c\u7684\u6700\u5c0f\u503c\u9009\u62e9\u5f80\u5f80\u4f1a\u5e26\u6765\u4e0d\u4e00\u6837\u7684\u62df\u5408\u7ed3\u679c\uff0c\u5982\u4e0b\u65b9\u4ee3\u7801\u6240\u793a\u3002<\/span><\/p>\n

<\/p>\n
\n
\n
>fit = powerlaw.Fit(data)<\/span>
>fit.xmin <\/span><\/span> #\u5f97\u5230\u6700\u4f18\u7684\u5e42\u5f8b\u62df\u5408\u6700\u5c0f\u503c<\/span><\/section>\n
230.000<\/span>
>fit.power_law.alpha
<\/span>2.273<\/span>
>fit.power_law.D <\/span> #\u9009\u62e9\u6570\u636e\u548c\u62df\u5408\u4e4b\u95f4\u7684Kolmogorov-Smirnov\u8ddd\u79bbD<\/span><\/section>\n
0.061<\/span>
>fit = powerlaw.Fit(data, xmin = <\/span>1.0<\/span>)  <\/span>#\u6307\u5b9ax_min\u8fdb\u884c\u5e42\u5f8b\u62df\u5408<\/span>
>fit.xmin
<\/span>1.0<\/span>
>fit.fixed_xmin
<\/span>True<\/span>
>fit.power_law.alpha
<\/span>1.220<\/span>
>fit.power_law.D
<\/span>0.376<\/span>
>fit = powerlaw.Fit(data, xmin = (<\/span>250.0<\/span>, <\/span>300.0<\/span>))  <\/span>#\u7ed9\u5b9ax_min\u8303\u56f4<\/span>
>fit.given_xmin<\/span>
(<\/span><\/span>250.000<\/span>, <\/span>300.000<\/span>)<\/span>
>fit.xmin<\/span>
<\/span>272.0<\/span><\/section>\n<\/section>\n<\/section>\n

<\/p>\n
\u53e6\u5916\uff0c\u4e00\u4e9b\u9886\u57df\u4e5f\u5e0c\u671b\u5206\u5e03\u6709\u4e00\u4e2a\u7cbe\u786e\u7684\u4e0a\u754cx_max\uff0c\u56e0\u4e3a\u6709\u65f6\u8003\u8651\u5b9e\u9645\u60c5\u51b5\u4e00\u4e9b\u6570\u636e\u53ef\u80fd\u8d85\u51fa\u4e86\u7406\u8bba\u6781\u9650<\/span>\uff08\u4f8b\u5982\u5728\u5929\u4f53\u7269\u7406\u5b66\u4e2d\uff0c\u901f\u5ea6\u7684\u5206\u5e03\u5728\u5149\u901f\u4e0b\u53ef\u80fd\u6709\u4e00\u4e2a\u4e0a\u9650\uff09<\/span>\u3002powerlaw\u5e93\u540c\u6837\u80fd\u6ee1\u8db3\u8fd9\u90e8\u5206\u7528\u6237\u7684\u9700\u6c42\u3002<\/span><\/p>\n

<\/p>\n
\n
\n
>fit = powerlaw.Fit(data, xmax = <\/span>10000.0<\/span>)<\/span>
>fit.xmax<\/span>
<\/span>10000.0<\/span>
>fit.fixed_xmax
<\/span>True<\/span><\/section>\n<\/section>\n<\/section>\n
       <\/span>\u56fe4. \u4e0d\u540c\u7684x_min\u548cx_max\u9009\u62e9\u53ef\u80fd\u5bfc\u81f4\u4e0d\u540c\u7684\u62df\u5408\u7ed3\u679c\u3002<\/span>
<\/p>\n

<\/p>\n
\u8fd9\u91cc\u7684x_min\u548cx_max\u9ed8\u8ba4\u4f7f\u7528KS\u8ddd\u79bbD\u6700\u5c0f\u7684\u503c\u3002\u7136\u800c\uff0c\u8fd9\u4e2a\u8ddd\u79bbD\u5bf9\u5206\u5e03\u5c3e\u90e8\u7684\u5dee\u5f02\u660e\u663e\u4e0d\u654f\u611f\uff0c\u800c\u5c3e\u90e8\u6b63\u662f\u5e42\u5f8b\u7684\u5927\u90e8\u5206\u6709\u8da3\u884c\u4e3a\u53d1\u751f\u7684\u5730\u65b9\u3002\u53ef\u80fd\u9700\u8981\u4f7f\u7528\u5176\u4ed6\u6307\u6807\uff0c\u5982Kuiper\u6216AndersonDarling\uff0c\u5b83\u4eec\u5728\u6d4b\u91cf\u5206\u5e03\u4e4b\u95f4\u7684\u8ddd\u79bb\u65f6\u7ed9\u4e88\u5c3e\u90e8\u989d\u5916\u7684\u6743\u91cd\uff0c\u7528\u6237\u53ef\u4ee5\u6839\u636e\u5b9e\u9645\u9700\u8981\u8fdb\u884c\u8c03\u6574\u3002<\/span><\/p>\n

<\/span><\/p>\n
>fit = powerlaw.Fit(data, xmin_distance = \u2018D\u2019)<\/span>
>fit = powerlaw.Fit(data, xmin_distance = \u2018V\u2019)<\/span>
>fit = powerlaw.Fit(data, xmin_distance = \u2018Asquare\u2019)<\/span><\/section><\/section><\/section><\/pre>\n

<\/span><\/strong><\/span><\/h4>\n
\n
\n
\n
\u79bb\u6563\u4e0e\u8fde\u7eed\u6570\u636e<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n

<\/p>\n<\/section>\n
\u4e3a\u4e86\u9002\u5408\u5e42\u5f8b\u548c\u5176\u4ed6\u5206\u5e03\u7684\u8fde\u7eed\u5f62\u5f0f\uff0c\u6570\u636e\u9ed8\u8ba4\u662f\u8fde\u7eed\u7684\u3002\u7136\u800c\uff0c\u8bb8\u591a\u6570\u636e\u662f\u79bb\u6563\u7684\u3002\u79bb\u6563\u5206\u5e03\u7684\u6982\u7387\u5206\u5e03\u4e0d\u80fd\u7cbe\u786e\u5730\u4e0e\u8fde\u7eed\u5206\u5e03\u76f8\u63d0\u5e76\u8bba\u3002\u79bb\u6563<\/span>\uff08\u6574\u6570\uff09<\/span>\u5206\u5e03\u901a\u8fc7\u9002\u5f53\u7684\u89c4\u8303\u5316\uff0c\u53ef\u4ee5\u5728\u521d\u59cb\u5316\u65f6\u6307\u5b9a\u3002\u4e0d\u8fc7\u79bb\u6563\u5f62\u5f0f\u7684\u6982\u7387\u5206\u5e03\u901a\u5e38\u6bd4\u8fde\u7eed\u5f62\u5f0f\u66f4\u96be\u8ba1\u7b97\uff0c\u56e0\u6b64\u67d0\u4e9b\u64cd\u4f5c\u8fd0\u884c\u53ef\u80fd\u4f1a\u6bd4\u8f83\u6162\u3002<\/span><\/p>\n

<\/p>\n
>fit = powerlaw.Fit(data, xmin = <\/span>230.0<\/span>)<\/span>
>fit.discrete<\/span>
<\/span>False<\/span>
>fit = powerlaw.Fit(data, xmin = <\/span>230.0<\/span>, discrete = <\/span>True<\/span>)<\/span>
>fit.discrete<\/span>
<\/span>True<\/span><\/section><\/section><\/section>

<\/p><\/pre>\n

<\/span><\/strong><\/h4>\n
\n
\n
\n
\u4e0e\u5176\u4ed6\u5206\u5e03\u6bd4\u8f83<\/span><\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n
\n

<\/p>\n<\/section>\n
\u4ece\u521b\u5efa\u7684Fit\u5bf9\u8c61\u4e2d\uff0c\u7528\u6237\u53ef\u4ee5\u5f88\u5bb9\u6613\u5730\u8bbf\u95ee\u8bc4\u4f30\u957f\u5c3e\u5206\u5e03\u6240\u9700\u7684\u6240\u6709\u7edf\u8ba1\u5206\u6790\u3002Fit\u5bf9\u8c61\u4e5f\u80fd\u9002\u7528\u4e8e\u5176\u4ed6\u53ef\u80fd\u7684\u5206\u5e03\u5f62\u5f0f\u3002\u6bcf\u4e2a\u5206\u5e03\u90fd\u6709\u9002\u5408\u8be5\u5206\u5e03\u7684\u6700\u4f73\u53c2\u6570\uff0c\u53ef\u901a\u8fc7\u53c2\u6570\u540d\u79f0\u6216\u66f4\u901a\u7528\u7684\u201cparameter1\u201d\u8bbf\u95ee\u3002<\/span><\/p>\n

<\/p>\n
\n
\n
>fit.power_law.alpha
<\/span>2.273<\/span>
>fit.power_law.parameter1
<\/span>2.273<\/span>
>fit.power_law.parameter1_name<\/span>
>fit.lognormal.mu<\/span>
<\/span>0.154<\/span>
>fit.lognormal.parameter1_name<\/span>
\u2018mu\u2019<\/span>
>fit.lognormal.parameter2_name<\/span>
\u2018sigma\u2019<\/span>
>fit.lognormal.parameter3_name = = <\/span><\/span>None<\/span><\/section>\n
True<\/span><\/section>\n<\/section>\n<\/section>\n

<\/p>\n
\u53e6\u5916\uff0c\u5f97\u5230\u53c2\u6570\u4e4b\u540e\u8fd8\u9700\u8981\u8bc4\u4ef7\u54ea\u4e2a\u5206\u5e03\u4f18\u5ea6\u66f4\u597d\u3002\u6bcf\u4e2a\u5206\u5e03\u7684\u62df\u5408\u4f18\u5ea6\u53ef\u4ee5\u8003\u8651\u5355\u72ec\u6216\u901a\u8fc7\u76f8\u4e92\u6bd4\u8f83\u5f97\u5230<\/span>\uff08\u8fd9\u91cc\u4f7f\u7528KS \u68c0\u9a8c\u548c\u5bf9\u6570\u4f3c\u7136\u6bd4\u6765\u786e\u5b9a\uff09<\/span>\u3002\u6bd4\u5982\u4e0b\u65b9\u4ee3\u7801\u76f4\u63a5\u5bf9\u6bd4\u5e42\u5f8b\u5206\u5e03\u548c\u6307\u6570\u5206\u5e03\u7684\u62df\u5408\u6548\u679c\u3002\u8fd4\u56de\u5f97\u5230\u7684R\u662f\u4e24\u4e2a\u5019\u9009\u5206\u5e03\u4e4b\u95f4\u7684\u5bf9\u6570\u4f3c\u7136\u6bd4\uff0c\u5982\u679c\u6570\u636e\u66f4\u6709\u53ef\u80fd\u5728\u7b2c\u4e00\u4e2a\u5206\u5e03\u4e2d\uff0c\u8fd9\u4e2a\u6570\u5b57\u5c06\u4e3a\u6b63\uff0c\u5982\u679c\u6570\u636e\u66f4\u6709\u53ef\u80fd\u5728\u7b2c\u4e8c\u4e2a\u5206\u5e03\u4e2d\uff0c\u8fd9\u4e2a\u6570\u5b57\u5c06\u4e3a\u8d1f\u3002p\u662f\u663e\u8457\u6027\u6307\u6807\u3002<\/span><\/p>\n

<\/span><\/p>\n
\n
\n
>R, p = fit.distribution_compare(\u2018power_law\u2019, \u2018exponential\u2019,<\/span>
normalized_ratio = <\/span><\/span>True<\/span>)<\/span>
><\/span><\/span>print<\/span> R, p
<\/span>16.384  0.024<\/span>  <\/span> #\u5e42\u5f8b\u5206\u5e03\u6bd4\u6307\u6570\u5206\u5e03\u66f4\u4f18<\/span><\/section>\n<\/section>\n<\/section>\n

<\/p>\n
\u4e00\u822c\u8ba4\u4e3a\uff0c\u6307\u6570\u5206\u5e03\u662f\u957f\u5c3e\u5206\u5e03\u4e2d\u7edd\u5bf9\u6700\u5c0f\u5907\u9009\u3002\u56e0\u6b64\uff0c\u5982\u679c\u5f53\u4e0b\u7ed3\u679c\u5e76\u4e0d\u6bd4\u6307\u6570\u5206\u5e03\u66f4\u9002\u5408\uff0c\u5c31\u6839\u672c\u6ca1\u6709\u7406\u7531\u8003\u8651\u5176\u4ed6\u7684\u957f\u5c3e\u5206\u5e03\uff0c\u66f4\u522b\u8bf4\u5e42\u5f8b\u5206\u5e03\u4e86\u3002fit\u5bf9\u8c61\u7684fit.supported_distribution\u4e2d\u652f\u6301\u7684\u5206\u5e03\u5217\u8868\u5305\u62ec\uff1a\u5e42\u5f8b\u3001\u622a\u65ad\u5e42\u5f8b\u3001\u6307\u6570\u3001\u5bf9\u6570\u6b63\u6001\u3001\u5a01\u5e03\u5c14\u548c\u4f3d\u9a6c\u5206\u5e03\u3002distribution_compare\u53ef\u4ee5\u4f7f\u7528\u8fd9\u4e9b\u5206\u5e03\u4e2d\u7684\u4efb\u4f55\u4e00\u4e2a\u3002<\/span><\/p>\n

<\/section>\n
\u53c2\u8003\u8d44\u6599\uff1a<\/span><\/strong><\/span><\/section>\n
1. \u8be5\u8bba\u6587\u4e2d\u7684\u62df\u5408\u6570\u636e\u548c\u4ee3\u7801\u5747\u53ef\u89c1\uff1ahttps:\/\/github.com\/jeffalstott\/powerlaw<\/span><\/section>\n
2. \u5173\u4e8e\u5e42\u5f8b\u5206\u5e03\u548c\u590d\u6742\u7f51\u7edc\u7684\u76f8\u5173\u8054\u7cfb\u53ef\u4ee5\u53c2\u8003\uff1a\u96c6\u667a\u4e4b\u524d\u5173\u4e8e\u5e42\u5f8b\u5206\u5e03\u7684\u4e00\u4e9b\u957f\u6587\u94fe\u63a5\u3002<\/span><\/section>\n
4. \u5173\u4e8e\u5e42\u5f8b\u5206\u5e03\u62df\u5408\u53ca\u68c0\u9a8c\u539f\u7406\u53ef\u4ee5\u53c2\u8003\u96c6\u667a\u5b66\u56ed\u9648\u6e05\u534e\u8001\u5e08\u6388\u8bfe\u89c6\u9891\uff1a<\/span><\/section>\n
https:\/\/campus.swarma.org\/course\/1747 \u3002<\/span><\/section>\n

<\/section>\n

<\/span><\/p>\n
\n
\n
\n
\n
\u8bfe\u7a0b\u63a8\u8350<\/strong><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n
\n

<\/p>\n<\/section>\n

<\/p>\n
\u8bfe\u7a0b\u5730\u5740\uff1ahttps:\/\/campus.swarma.org\/course\/647<\/strong><\/span><\/p>\n
\n

<\/p>\n<\/section>\n

<\/p>\n
\n
\u63a8\u8350\u9605\u8bfb<\/span><\/strong><\/strong><\/p>\n