{"id":985956,"date":"2024-12-27T07:40:55","date_gmt":"2024-12-26T23:40:55","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/985956.html"},"modified":"2024-12-27T07:40:57","modified_gmt":"2024-12-26T23:40:57","slug":"python%e5%a6%82%e4%bd%95%e7%94%bbks%e6%9b%b2%e7%ba%bf","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/985956.html","title":{"rendered":"python\u5982\u4f55\u753bks\u66f2\u7ebf"},"content":{"rendered":"

\"python\u5982\u4f55\u753bks\u66f2\u7ebf\"<\/p>\n

\u5728Python\u4e2d\uff0c\u753bKS\uff08Kolmogorov-Smirnov\uff09\u66f2\u7ebf\u7684\u8fc7\u7a0b\u4e3b\u8981\u6d89\u53ca\u5230\u6570\u636e\u7684\u51c6\u5907\u3001\u6a21\u578b\u9884\u6d4b\u3001\u8ba1\u7b97KS\u7edf\u8ba1\u91cf\u4ee5\u53ca\u7ed8\u5236\u66f2\u7ebf\u3002Python\u753bKS\u66f2\u7ebf\u7684\u6838\u5fc3\u6b65\u9aa4\u5305\u62ec\uff1a\u6570\u636e\u51c6\u5907\u3001\u6a21\u578b\u9884\u6d4b\u3001\u8ba1\u7b97KS\u7edf\u8ba1\u91cf\u3001\u7ed8\u5236KS\u66f2\u7ebf<\/strong>\u3002\u5176\u4e2d\uff0c\u8ba1\u7b97KS\u7edf\u8ba1\u91cf\u662f\u5173\u952e\u6b65\u9aa4\uff0c\u5b83\u8861\u91cf\u4e86\u6a21\u578b\u533a\u5206\u6b63\u8d1f\u6837\u672c\u7684\u80fd\u529b\u3002\u4e0b\u9762\uff0c\u6211\u5c06\u8be6\u7ec6\u4ecb\u7ecd\u5982\u4f55\u5728Python\u4e2d\u7ed8\u5236KS\u66f2\u7ebf\u3002<\/p>\n<\/p>\n

\u4e00\u3001\u6570\u636e\u51c6\u5907<\/h3>\n<\/p>\n

\u5728\u7ed8\u5236KS\u66f2\u7ebf\u4e4b\u524d\uff0c\u9996\u5148\u9700\u8981\u51c6\u5907\u597d\u6570\u636e\u3002\u901a\u5e38\u60c5\u51b5\u4e0b\uff0c\u6570\u636e\u96c6\u5e94\u5305\u542b\u76ee\u6807\u53d8\u91cf\uff08\u901a\u5e38\u662f\u4e8c\u5206\u7c7b\u7684\uff09\u548c\u7279\u5f81\u53d8\u91cf\u3002\u76ee\u6807\u53d8\u91cf\u7528\u4e8e\u6307\u793a\u6b63\u8d1f\u6837\u672c\uff0c\u800c\u7279\u5f81\u53d8\u91cf\u7528\u4e8e\u9884\u6d4b\u3002<\/p>\n<\/p>\n

    \n
  1. \n

    \u52a0\u8f7d\u6570\u636e\u96c6<\/strong><\/p>\n<\/p>\n

    \u53ef\u4ee5\u4f7f\u7528pandas\u5e93\u52a0\u8f7d\u6570\u636e\u96c6\uff0c\u5e38\u89c1\u7684\u6570\u636e\u683c\u5f0f\u5305\u62ecCSV\u3001Excel\u7b49\u3002\u4ee5\u4e0b\u662f\u4f7f\u7528pandas\u52a0\u8f7dCSV\u6587\u4ef6\u7684\u793a\u4f8b\uff1a<\/p>\n<\/p>\n

    import pandas as pd<\/p>\n
    \u52a0\u8f7d\u6570\u636e<\/strong><\/h2>\n
    data = pd.read_csv('your_dataset.csv')<\/p>\n
    <\/code><\/pre>\n<\/p>\n<\/li>\n
  2. \n
    \u6570\u636e\u9884\u5904\u7406<\/strong><\/p>\n<\/p>\n
    \u5728\u6570\u636e\u52a0\u8f7d\u540e\uff0c\u901a\u5e38\u9700\u8981\u8fdb\u884c\u4e00\u4e9b\u9884\u5904\u7406\u6b65\u9aa4\uff0c\u4f8b\u5982\u7f3a\u5931\u503c\u5904\u7406\u3001\u6570\u636e\u6e05\u6d17\u548c\u7279\u5f81\u5de5\u7a0b\u7b49\u3002\u8fd9\u4e9b\u6b65\u9aa4\u53ef\u4ee5\u5e2e\u52a9\u63d0\u5347\u6a21\u578b\u6027\u80fd\u3002<\/p>\n<\/p>\n<\/li>\n<\/ol>\n
    \u4e8c\u3001\u6a21\u578b\u9884\u6d4b<\/h3>\n<\/p>\n
    \u5728\u6570\u636e\u51c6\u5907\u5b8c\u6bd5\u540e\uff0c\u4e0b\u4e00\u6b65\u662f\u4f7f\u7528\u673a\u5668\u5b66\u4e60<\/a>\u6a21\u578b\u8fdb\u884c\u9884\u6d4b\u3002\u5e38\u7528\u7684\u6a21\u578b\u5305\u62ec\u903b\u8f91\u56de\u5f52\u3001\u51b3\u7b56\u6811\u3001\u968f\u673a\u68ee\u6797\u7b49\u3002\u9009\u62e9\u5408\u9002\u7684\u6a21\u578b\u53d6\u51b3\u4e8e\u5177\u4f53\u7684\u4e1a\u52a1\u9700\u6c42\u548c\u6570\u636e\u7279\u6027\u3002<\/p>\n<\/p>\n
      \n
    1. \n
      \u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6<\/strong><\/p>\n<\/p>\n
      \u901a\u5e38\u60c5\u51b5\u4e0b\uff0c\u6211\u4eec\u4f1a\u5c06\u6570\u636e\u96c6\u5212\u5206\u4e3a\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\uff0c\u4ee5\u4fbf\u8bc4\u4f30\u6a21\u578b\u6027\u80fd\u3002\u53ef\u4ee5\u4f7f\u7528scikit-learn\u5e93\u4e2d\u7684trAI<\/a>n_test_split<\/code>\u51fd\u6570\u6765\u5b9e\u73b0\uff1a<\/p>\n<\/p>\n
      from sklearn.model_selection import train_test_split<\/p>\n
      \u7279\u5f81\u548c\u76ee\u6807\u53d8\u91cf<\/strong><\/h2>\n
      X = data.drop('target', axis=1)<\/p>\n
      y = data['target']<\/p>\n
      \u5212\u5206\u6570\u636e\u96c6<\/strong><\/h2>\n
      X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)<\/p>\n
      <\/code><\/pre>\n<\/p>\n<\/li>\n
    2. \n
      \u8bad\u7ec3\u6a21\u578b<\/strong><\/p>\n<\/p>\n
      \u4ee5\u903b\u8f91\u56de\u5f52\u4e3a\u4f8b\uff0c\u8bad\u7ec3\u6a21\u578b\u5e76\u8fdb\u884c\u9884\u6d4b\uff1a<\/p>\n<\/p>\n
      from sklearn.linear_model import LogisticRegression<\/p>\n
      \u521d\u59cb\u5316\u6a21\u578b<\/strong><\/h2>\n
      model = LogisticRegression()<\/p>\n
      \u8bad\u7ec3\u6a21\u578b<\/strong><\/h2>\n
      model.fit(X_train, y_train)<\/p>\n
      \u6a21\u578b\u9884\u6d4b<\/strong><\/h2>\n
      y_pred_prob = model.predict_proba(X_test)[:, 1]<\/p>\n
      <\/code><\/pre>\n<\/p>\n<\/li>\n<\/ol>\n
      \u4e09\u3001\u8ba1\u7b97KS\u7edf\u8ba1\u91cf<\/h3>\n<\/p>\n
      \u8ba1\u7b97KS\u7edf\u8ba1\u91cf\u662f\u7ed8\u5236KS\u66f2\u7ebf\u7684\u91cd\u8981\u6b65\u9aa4\u3002KS\u7edf\u8ba1\u91cf\u8861\u91cf\u4e86\u6a21\u578b\u5bf9\u6b63\u8d1f\u6837\u672c\u7684\u533a\u5206\u80fd\u529b\uff0c\u901a\u5e38\u7528\u4e8e\u8bc4\u4f30\u4e8c\u5206\u7c7b\u6a21\u578b\u7684\u6027\u80fd\u3002<\/p>\n<\/p>\n
        \n
      1. \n
        \u8ba1\u7b97KS\u503c<\/strong><\/p>\n<\/p>\n
        \u53ef\u4ee5\u4f7f\u7528scipy\u5e93\u7684ks_2samp<\/code>\u51fd\u6570\u6765\u8ba1\u7b97KS\u7edf\u8ba1\u91cf\uff1a<\/p>\n<\/p>\n
        from scipy.stats import ks_2samp<\/p>\n
        \u8ba1\u7b97KS\u503c<\/strong><\/h2>\n
        ks_statistic, p_value = ks_2samp(y_test[y_pred_prob > 0.5], y_test[y_pred_prob <= 0.5])<\/p>\n
        print(f'KS Statistic: {ks_statistic}')<\/p>\n
        <\/code><\/pre>\n<\/p>\n<\/li>\n
      2. \n
        \u7ed8\u5236\u7d2f\u79ef\u5206\u5e03\u51fd\u6570\uff08CDF\uff09<\/strong><\/p>\n<\/p>\n
        \u5728\u7ed8\u5236KS\u66f2\u7ebf\u65f6\uff0c\u901a\u5e38\u4f1a\u8ba1\u7b97\u6b63\u8d1f\u6837\u672c\u7684\u7d2f\u79ef\u5206\u5e03\u51fd\u6570\uff08CDF\uff09\u3002\u53ef\u4ee5\u4f7f\u7528numpy\u5e93\u5b9e\u73b0\uff1a<\/p>\n<\/p>\n
        import numpy as np<\/p>\n
        \u8ba1\u7b97CDF<\/strong><\/h2>\n
        pos_cdf = np.cumsum(np.sort(y_pred_prob[y_test == 1]))<\/p>\n
        neg_cdf = np.cumsum(np.sort(y_pred_prob[y_test == 0]))<\/p>\n
        <\/code><\/pre>\n<\/p>\n<\/li>\n<\/ol>\n
        \u56db\u3001\u7ed8\u5236KS\u66f2\u7ebf<\/h3>\n<\/p>\n
        \u5728\u5b8c\u6210\u4e0a\u8ff0\u6b65\u9aa4\u540e\uff0c\u53ef\u4ee5\u4f7f\u7528matplotlib\u5e93\u7ed8\u5236KS\u66f2\u7ebf\u3002<\/p>\n<\/p>\n
          \n
        1. \n
          \u7ed8\u5236\u66f2\u7ebf<\/strong><\/p>\n<\/p>\n
          \u4f7f\u7528matplotlib\u7ed8\u5236\u6b63\u8d1f\u6837\u672c\u7684CDF\u66f2\u7ebf\uff0c\u5e76\u6807\u8bb0KS\u7edf\u8ba1\u91cf\uff1a<\/p>\n<\/p>\n
          import matplotlib.pyplot as plt<\/p>\n
          \u7ed8\u5236KS\u66f2\u7ebf<\/strong><\/h2>\n
          plt.plot(pos_cdf, label='Positive CDF')<\/p>\n
          plt.plot(neg_cdf, label='Negative CDF')<\/p>\n
          plt.title('KS Curve')<\/p>\n
          plt.xlabel('Sample Index')<\/p>\n
          plt.ylabel('CDF')<\/p>\n
          plt.legend(loc='best')<\/p>\n
          \u6807\u8bb0KS\u7edf\u8ba1\u91cf<\/strong><\/h2>\n
          plt.axvline(x=ks_statistic, color='r', linestyle='--', label=f'KS Statistic = {ks_statistic:.2f}')<\/p>\n
          plt.legend()<\/p>\n
          plt.show()<\/p>\n
          <\/code><\/pre>\n<\/p>\n<\/li>\n
        2. \n
          \u89e3\u91caKS\u66f2\u7ebf<\/strong><\/p>\n<\/p>\n
          KS\u66f2\u7ebf\u53cd\u6620\u4e86\u6a21\u578b\u5bf9\u6b63\u8d1f\u6837\u672c\u7684\u533a\u5206\u80fd\u529b\u3002\u66f2\u7ebf\u4e4b\u95f4\u7684\u6700\u5927\u5782\u76f4\u8ddd\u79bb\u5373\u4e3aKS\u7edf\u8ba1\u91cf\uff0c\u901a\u5e38\u8be5\u503c\u8d8a\u5927\uff0c\u6a21\u578b\u6027\u80fd\u8d8a\u597d\u3002\u901a\u8fc7\u89c2\u5bdf\u66f2\u7ebf\uff0c\u53ef\u4ee5\u76f4\u89c2\u5730\u8bc4\u4f30\u6a21\u578b\u7684\u5206\u7c7b\u6548\u679c\u3002<\/p>\n<\/p>\n<\/li>\n<\/ol>\n
          \u4e94\u3001\u4f18\u5316\u4e0e\u8c03\u4f18<\/h3>\n<\/p>\n
          \u5728\u7ed8\u5236\u5b8cKS\u66f2\u7ebf\u540e\uff0c\u53ef\u4ee5\u6839\u636e\u66f2\u7ebf\u7ed3\u679c\u8fdb\u4e00\u6b65\u4f18\u5316\u6a21\u578b\u3002\u4f8b\u5982\uff0c\u901a\u8fc7\u8c03\u6574\u6a21\u578b\u53c2\u6570\u3001\u8fdb\u884c\u7279\u5f81\u9009\u62e9\u6216\u4f7f\u7528\u66f4\u590d\u6742\u7684\u6a21\u578b\u6765\u63d0\u5347\u6027\u80fd\u3002\u6b64\u5916\uff0c\u53ef\u4ee5\u8fdb\u884c\u4ea4\u53c9\u9a8c\u8bc1\u4ee5\u83b7\u5f97\u66f4\u7a33\u5065\u7684\u6027\u80fd\u8bc4\u4f30\u3002<\/p>\n<\/p>\n
          \u516d\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
          \u7ed8\u5236KS\u66f2\u7ebf\u662f\u8bc4\u4f30\u4e8c\u5206\u7c7b\u6a21\u578b\u6027\u80fd\u7684\u91cd\u8981\u6b65\u9aa4\u3002\u901a\u8fc7\u4e0a\u8ff0\u6b65\u9aa4\uff0c\u53ef\u4ee5\u5728Python\u4e2d\u9ad8\u6548\u5730\u7ed8\u5236KS\u66f2\u7ebf\uff0c\u5e76\u6839\u636e\u66f2\u7ebf\u7ed3\u679c\u4f18\u5316\u6a21\u578b\u6027\u80fd\u3002\u65e0\u8bba\u662f\u6570\u636e\u51c6\u5907\u3001\u6a21\u578b\u9884\u6d4b\u8fd8\u662f\u66f2\u7ebf\u7ed8\u5236\uff0c\u6bcf\u4e00\u6b65\u90fd\u81f3\u5173\u91cd\u8981\u3002\u901a\u8fc7\u4e0d\u65ad\u4f18\u5316\u5404\u4e2a\u73af\u8282\uff0c\u53ef\u4ee5\u6709\u6548\u63d0\u5347\u6a21\u578b\u7684\u533a\u5206\u80fd\u529b\uff0c\u4ece\u800c\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\u53d6\u5f97\u66f4\u597d\u7684\u6548\u679c\u3002<\/p>\n<\/p>\n
          \u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
           \u5982\u4f55\u4f7f\u7528Python\u7ed8\u5236KS\u66f2\u7ebf\uff1f<\/strong>
          \u5728Python\u4e2d\uff0c\u7ed8\u5236KS\u66f2\u7ebf\u901a\u5e38\u53ef\u4ee5\u901a\u8fc7\u4f7f\u7528SciPy\u548cMatplotlib\u5e93\u6765\u5b9e\u73b0\u3002\u9996\u5148\uff0c\u4f60\u9700\u8981\u8ba1\u7b97\u7ecf\u9a8c\u5206\u5e03\u51fd\u6570\uff08CDF\uff09\u5e76\u7ed8\u5236\u5b9e\u9645\u503c\u4e0e\u9884\u6d4b\u503c\u7684CDF\u3002\u7136\u540e\uff0c\u4f7f\u7528Matplotlib\u7ed8\u5236\u4e24\u6761\u66f2\u7ebf\u7684\u5dee\u5f02\uff0c\u5373KS\u7edf\u8ba1\u91cf\u3002\u53ef\u4ee5\u53c2\u8003\u4ee5\u4e0b\u4ee3\u7801\u793a\u4f8b\uff1a  <\/p>\n
          import numpy as np\nimport matplotlib.pyplot as plt\nfrom scipy import stats\n\n# \u751f\u6210\u793a\u4f8b\u6570\u636e\ndata1 = np.random.normal(0, 1, 1000)\ndata2 = np.random.normal(0, 1.5, 1000)\n\n# \u8ba1\u7b97CDF\necdf1 = np.sort(data1)\necdf2 = np.sort(data2)\ncdf1 = np.arange(1, len(ecdf1) + 1) \/ len(ecdf1)\ncdf2 = np.arange(1, len(ecdf2) + 1) \/ len(ecdf2)\n\n# \u7ed8\u5236KS\u66f2\u7ebf\nplt.step(ecdf1, cdf1, label='Sample 1 CDF', where='post')\nplt.step(ecdf2, cdf2, label='Sample 2 CDF', where='post')\nplt.title('KS Curve')\nplt.xlabel('Value')\nplt.ylabel('Cumulative Probability')\nplt.legend()\nplt.grid()\nplt.show()\n<\/code><\/pre>\n
          KS\u66f2\u7ebf\u7684\u5177\u4f53\u5e94\u7528\u573a\u666f\u6709\u54ea\u4e9b\uff1f<\/strong>
          KS\u66f2\u7ebf\u5e7f\u6cdb\u5e94\u7528\u4e8e\u7edf\u8ba1\u5206\u6790\u548c\u673a\u5668\u5b66\u4e60\u6a21\u578b\u7684\u6027\u80fd\u8bc4\u4f30\u4e2d\u3002\u5e38\u89c1\u7684\u5e94\u7528\u573a\u666f\u5305\u62ec\u4fe1\u7528\u8bc4\u5206\u6a21\u578b\u7684\u9a8c\u8bc1\u3001\u4e8c\u5143\u5206\u7c7b\u6a21\u578b\u7684\u6548\u679c\u5206\u6790\uff0c\u4ee5\u53ca\u4efb\u4f55\u9700\u8981\u6bd4\u8f83\u4e24\u4e2a\u5206\u5e03\u7684\u573a\u5408\u3002\u901a\u8fc7KS\u66f2\u7ebf\uff0c\u7814\u7a76\u8005\u80fd\u591f\u76f4\u89c2\u4e86\u89e3\u6a21\u578b\u7684\u533a\u5206\u80fd\u529b\uff0c\u4ece\u800c\u8fdb\u884c\u4f18\u5316\u8c03\u6574\u3002<\/p>\n
          \u5728Python\u4e2d\u7ed8\u5236KS\u66f2\u7ebf\u9700\u8981\u54ea\u4e9b\u5e93\uff1f<\/strong>
          \u7ed8\u5236KS\u66f2\u7ebf\u901a\u5e38\u9700\u8981\u5b89\u88c5\u51e0\u4e2a\u6838\u5fc3\u5e93\u3002\u6700\u5e38\u7528\u7684\u5305\u62ecNumPy\uff08\u7528\u4e8e\u6570\u636e\u5904\u7406\uff09\u3001Matplotlib\uff08\u7528\u4e8e\u7ed8\u56fe\uff09\u4ee5\u53caSciPy\uff08\u7528\u4e8e\u79d1\u5b66\u8ba1\u7b97\u548c\u7edf\u8ba1\u5206\u6790\uff09\u3002\u786e\u4fdd\u5728\u4f60\u7684Python\u73af\u5883\u4e2d\u5b89\u88c5\u4e86\u8fd9\u4e9b\u5e93\uff0c\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u547d\u4ee4\u8fdb\u884c\u5b89\u88c5\uff1a  <\/p>\n
          pip install numpy matplotlib scipy\n<\/code><\/pre>\n
          \u5b89\u88c5\u5b8c\u6210\u540e\uff0c\u5c31\u53ef\u4ee5\u5f00\u59cb\u7ed8\u5236KS\u66f2\u7ebf\u4e86\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u5728Python\u4e2d\uff0c\u753bKS\uff08Kolmogorov-Smirnov\uff09\u66f2\u7ebf\u7684\u8fc7\u7a0b\u4e3b\u8981\u6d89\u53ca\u5230\u6570\u636e\u7684\u51c6\u5907\u3001\u6a21\u578b\u9884\u6d4b\u3001\u8ba1\u7b97 […]","protected":false},"author":3,"featured_media":985960,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/985956"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=985956"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/985956\/revisions"}],"predecessor-version":[{"id":985966,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/985956\/revisions\/985966"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/985960"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=985956"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=985956"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=985956"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}