\u4e8c\u7ef4\u53ef\u89c6\u5316<\/strong><\/p>\n<\/p>\n\u5f53\u6211\u4eec\u4fdd\u7559\u4e24\u4e2a\u4e3b\u6210\u5206\u65f6\uff0c\u53ef\u4ee5\u5c06\u5176\u7ed8\u5236\u6210\u4e8c\u7ef4\u56fe\u5f62\u6765\u67e5\u770b\u6570\u636e\u7684\u5206\u5e03\u3002<\/p>\n<\/p>\n
import matplotlib.pyplot as plt<\/p>\nimport seaborn as sns<\/p>\n
\u7ed8\u5236\u4e3b\u6210\u5206\u7684\u6563\u70b9\u56fe<\/strong><\/h2>\nplt.figure(figsize=(8,6))<\/p>\n
sns.scatterplot(x='principal component 1', y='principal component 2', data=df_pca)<\/p>\n
plt.title('2D PCA')<\/p>\n
plt.xlabel('Principal Component 1')<\/p>\n
plt.ylabel('Principal Component 2')<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n
\n\u4e09\u7ef4\u53ef\u89c6\u5316<\/strong><\/p>\n<\/p>\n\u5982\u679c\u4fdd\u7559\u4e86\u4e09\u4e2a\u4e3b\u6210\u5206\uff0c\u53ef\u4ee5\u4f7f\u75283D\u56fe\u6765\u67e5\u770b\u6570\u636e\u7684\u5206\u5e03\u3002<\/p>\n<\/p>\n
from mpl_toolkits.mplot3d import Axes3D<\/p>\nfig = plt.figure(figsize=(8,6))<\/p>\n
ax = fig.add_subplot(111, projection='3d')<\/p>\n
ax.scatter(df_pca['principal component 1'], df_pca['principal component 2'], df_pca['principal component 3'])<\/p>\n
ax.set_title('3D PCA')<\/p>\n
ax.set_xlabel('Principal Component 1')<\/p>\n
ax.set_ylabel('Principal Component 2')<\/p>\n
ax.set_zlabel('Principal Component 3')<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n<\/ol>\n
\u56db\u3001PCA\u5e94\u7528\u5b9e\u4f8b<\/h3>\n<\/p>\n
\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\uff0cPCA\u5e38\u7528\u4e8e\u6570\u636e\u964d\u7ef4\u3001\u7279\u5f81\u9009\u62e9\u548c\u6570\u636e\u53ef\u89c6\u5316\u3002\u4ee5\u4e0b\u662f\u4e00\u4e9b\u5e38\u89c1\u7684PCA\u5e94\u7528\u5b9e\u4f8b\u3002<\/p>\n<\/p>\n
\n- \n
\u56fe\u50cf\u538b\u7f29<\/strong><\/p>\n<\/p>\n\u5728\u56fe\u50cf\u5904\u7406\u4e2d\uff0cPCA\u53ef\u4ee5\u7528\u6765\u964d\u4f4e\u56fe\u50cf\u7684\u7ef4\u5ea6\uff0c\u4ece\u800c\u5b9e\u73b0\u538b\u7f29\u3002\u901a\u8fc7\u4fdd\u7559\u4e3b\u8981\u7684\u4e3b\u6210\u5206\uff0c\u6211\u4eec\u53ef\u4ee5\u5728\u4e0d\u660e\u663e\u635f\u5931\u56fe\u50cf\u8d28\u91cf\u7684\u60c5\u51b5\u4e0b\u51cf\u5c11\u5b58\u50a8\u7a7a\u95f4\u3002<\/p>\n<\/p>\n
from sklearn.datasets import load_digits<\/p>\n\u52a0\u8f7d\u624b\u5199\u6570\u5b57\u6570\u636e\u96c6<\/strong><\/h2>\ndigits = load_digits()<\/p>\n
\u6807\u51c6\u5316\u6570\u636e<\/strong><\/h2>\ndata = StandardScaler().fit_transform(digits.data)<\/p>\n
\u6267\u884cPCA<\/strong><\/h2>\npca = PCA(n_components=0.95)<\/p>\n
reduced_data = pca.fit_transform(data)<\/p>\n
\u67e5\u770b\u538b\u7f29\u540e\u7684\u6570\u636e<\/strong><\/h2>\nprint('Reduced data shape:', reduced_data.shape)<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n
- \n
\u5e02\u573a\u5206\u6790<\/strong><\/p>\n<\/p>\n\u5728\u5e02\u573a\u5206\u6790\u4e2d\uff0cPCA\u53ef\u4ee5\u7528\u4e8e\u8bc6\u522b\u548c\u53bb\u9664\u5197\u4f59\u4fe1\u606f\uff0c\u5e2e\u52a9\u5206\u6790\u4eba\u5458\u66f4\u597d\u5730\u7406\u89e3\u4e0d\u540c\u5e02\u573a\u56e0\u7d20\u4e4b\u95f4\u7684\u5173\u7cfb\u3002<\/p>\n<\/p>\n
# \u5047\u8bbe\u6211\u4eec\u6709\u5e02\u573a\u6570\u636e<\/p>\nmarket_data = pd.read_csv('market_data.csv')<\/p>\n
\u6807\u51c6\u5316\u5e02\u573a\u6570\u636e<\/strong><\/h2>\nmarket_data_standardized = StandardScaler().fit_transform(market_data)<\/p>\n
\u6267\u884cPCA<\/strong><\/h2>\npca = PCA(n_components=3)<\/p>\n
market_pca = pca.fit_transform(market_data_standardized)<\/p>\n
\u67e5\u770b\u4e3b\u6210\u5206\u5206\u6790\u540e\u7684\u5e02\u573a\u6570\u636e<\/strong><\/h2>\nprint(pd.DataFrame(market_pca, columns=['PC1', 'PC2', 'PC3']).head())<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n
- \n
\u57fa\u56e0\u8868\u8fbe\u6570\u636e\u5206\u6790<\/strong><\/p>\n<\/p>\n\u5728\u751f\u7269\u4fe1\u606f\u5b66\u4e2d\uff0cPCA\u88ab\u5e7f\u6cdb\u5e94\u7528\u4e8e\u57fa\u56e0\u8868\u8fbe\u6570\u636e\u7684\u5206\u6790\uff0c\u4ee5\u8bc6\u522b\u4e0d\u540c\u57fa\u56e0\u7684\u8868\u8fbe\u6a21\u5f0f\u3002<\/p>\n<\/p>\n
# \u5047\u8bbe\u6211\u4eec\u6709\u57fa\u56e0\u8868\u8fbe\u6570\u636e<\/p>\ngene_data = pd.read_csv('gene_expression.csv')<\/p>\n
\u6807\u51c6\u5316\u57fa\u56e0\u6570\u636e<\/strong><\/h2>\ngene_data_standardized = StandardScaler().fit_transform(gene_data)<\/p>\n
\u6267\u884cPCA<\/strong><\/h2>\npca = PCA(n_components=5)<\/p>\n
gene_pca = pca.fit_transform(gene_data_standardized)<\/p>\n
\u67e5\u770b\u4e3b\u6210\u5206\u5206\u6790\u540e\u7684\u57fa\u56e0\u8868\u8fbe\u6570\u636e<\/strong><\/h2>\nprint(pd.DataFrame(gene_pca, columns=['PC1', 'PC2', 'PC3', 'PC4', 'PC5']).head())<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n<\/ol>\n
\u4e94\u3001PCA\u7684\u5c40\u9650\u6027\u4e0e\u4f18\u5316<\/h3>\n<\/p>\n
\u5c3d\u7ba1PCA\u662f\u4e00\u4e2a\u5f3a\u5927\u7684\u5de5\u5177\uff0c\u4f46\u5b83\u4e5f\u6709\u5176\u5c40\u9650\u6027\u3002\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\uff0c\u6211\u4eec\u9700\u8981\u6ce8\u610f\u8fd9\u4e9b\u9650\u5236\uff0c\u5e76\u8003\u8651\u53ef\u80fd\u7684\u4f18\u5316\u7b56\u7565\u3002<\/p>\n<\/p>\n
\n- \n
\u7ebf\u6027\u5047\u8bbe<\/strong><\/p>\n<\/p>\nPCA\u5047\u8bbe\u6570\u636e\u662f\u7ebf\u6027\u53ef\u5206\u7684\uff0c\u8fd9\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\u53ef\u80fd\u4e0d\u6210\u7acb\u3002\u5bf9\u4e8e\u975e\u7ebf\u6027\u6570\u636e\uff0c\u53ef\u80fd\u9700\u8981\u4f7f\u7528\u975e\u7ebf\u6027\u964d\u7ef4\u65b9\u6cd5\uff0c\u5982t-SNE\u6216UMAP\u3002<\/p>\n<\/p>\n
from sklearn.manifold import TSNE<\/p>\n\u4f7f\u7528t-SNE\u8fdb\u884c\u975e\u7ebf\u6027\u964d\u7ef4<\/strong><\/h2>\ntsne = TSNE(n_components=2)<\/p>\n
tsne_results = tsne.fit_transform(df_standardized)<\/p>\n
\u7ed8\u5236t-SNE\u7ed3\u679c<\/strong><\/h2>\nplt.figure(figsize=(8,6))<\/p>\n
sns.scatterplot(x=tsne_results[:,0], y=tsne_results[:,1])<\/p>\n
plt.title('t-SNE')<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n
- \n
\u7279\u5f81\u7f29\u653e\u7684\u91cd\u8981\u6027<\/strong><\/p>\n<\/p>\n\u5728\u67d0\u4e9b\u60c5\u51b5\u4e0b\uff0c\u7279\u5f81\u7684\u7f29\u653e\u53ef\u80fd\u4f1a\u5bf9PCA\u7ed3\u679c\u4ea7\u751f\u91cd\u5927\u5f71\u54cd\u3002\u56e0\u6b64\uff0c\u9009\u62e9\u5408\u9002\u7684\u6807\u51c6\u5316\u65b9\u6cd5\u81f3\u5173\u91cd\u8981\u3002<\/p>\n<\/p>\n
from sklearn.preprocessing import MinMaxScaler<\/p>\n\u4f7f\u7528MinMaxScaler\u8fdb\u884c\u6570\u636e\u7f29\u653e<\/strong><\/h2>\nmin_max_scaler = MinMaxScaler()<\/p>\n
scaled_data = min_max_scaler.fit_transform(df.loc[:, features].values)<\/p>\n
\u6267\u884cPCA<\/strong><\/h2>\npca = PCA(n_components=2)<\/p>\n
scaled_pca = pca.fit_transform(scaled_data)<\/p>\n
\u67e5\u770b\u7f29\u653e\u540e\u7684PCA\u7ed3\u679c<\/strong><\/h2>\nprint(pd.DataFrame(scaled_pca, columns=['PC1', 'PC2']).head())<\/p>\n
<\/code><\/pre>\n<\/p>\n<\/li>\n
- \n
\u89e3\u91ca\u6027\u7684\u9650\u5236<\/strong><\/p>\n<\/p>\n\u867d\u7136PCA\u63d0\u4f9b\u4e86\u964d\u7ef4\u548c\u53bb\u566a\u7684\u80fd\u529b\uff0c\u4f46\u4e3b\u6210\u5206\u672c\u8eab\u5f80\u5f80\u96be\u4ee5\u89e3\u91ca\u3002\u56e0\u6b64\uff0c\u5728\u5e94\u7528PCA\u65f6\uff0c\u7ed3\u5408\u9886\u57df\u77e5\u8bc6\u5bf9\u7ed3\u679c\u8fdb\u884c\u89e3\u91ca\u662f\u81f3\u5173\u91cd\u8981\u7684\u3002<\/p>\n<\/p>\n
\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u8be6\u7ec6\u4ecb\u7ecd\u4e86\u5982\u4f55\u5728Python\u4e2d\u4f7f\u7528PCA\u8fdb\u884c\u6570\u636e\u5206\u6790\uff0c\u5305\u62ec\u6570\u636e\u6807\u51c6\u5316\u3001PCA\u6267\u884c\u3001\u7ed3\u679c\u53ef\u89c6\u5316\u4ee5\u53caPCA\u7684\u5e94\u7528\u5b9e\u4f8b\u548c\u5c40\u9650\u6027\u3002\u901a\u8fc7\u8fd9\u79cd\u65b9\u6cd5\uff0c\u60a8\u53ef\u4ee5\u66f4\u597d\u5730\u7406\u89e3\u6570\u636e\u7684\u5185\u5728\u7ed3\u6784\uff0c\u5e76\u5728\u5404\u79cd\u5b9e\u9645\u5e94\u7528\u4e2d\u5229\u7528PCA\u7684\u5f3a\u5927\u529f\u80fd\u3002<\/p>\n<\/p>\n<\/li>\n<\/ol>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n \u4ec0\u4e48\u662fPCA\uff0c\u4e3a\u4ec0\u4e48\u5728Python\u4e2d\u4f7f\u7528\u5b83\uff1f<\/strong>
\u4e3b\u6210\u5206\u5206\u6790\uff08PCA\uff09\u662f\u4e00\u79cd\u5e38\u7528\u7684\u6570\u636e\u964d\u7ef4\u6280\u672f\uff0c\u5b83\u80fd\u591f\u901a\u8fc7\u7ebf\u6027\u53d8\u6362\u5c06\u6570\u636e\u4ece\u9ad8\u7ef4\u7a7a\u95f4\u6295\u5f71\u5230\u4f4e\u7ef4\u7a7a\u95f4\uff0c\u4ece\u800c\u4fdd\u7559\u5c3d\u53ef\u80fd\u591a\u7684\u539f\u59cb\u6570\u636e\u65b9\u5dee\u3002\u5728Python\u4e2d\uff0cPCA\u901a\u5e38\u7528\u4e8e\u6570\u636e\u9884\u5904\u7406\u3001\u7279\u5f81\u63d0\u53d6\u548c\u53ef\u89c6\u5316\uff0c\u7279\u522b\u662f\u5728\u5904\u7406\u5927\u89c4\u6a21\u6570\u636e\u96c6\u65f6\u975e\u5e38\u6709\u6548\u3002<\/p>\n\u5728Python\u4e2d\u5982\u4f55\u5b9e\u73b0PCA\uff1f<\/strong>
\u5728Python\u4e2d\uff0cPCA\u53ef\u4ee5\u901a\u8fc7\u591a\u79cd\u5e93\u5b9e\u73b0\uff0c\u6700\u5e38\u7528\u7684\u662fscikit-learn<\/code>\u3002\u4f7f\u7528PCA<\/code>\u7c7b\uff0c\u7528\u6237\u53ef\u4ee5\u8f7b\u677e\u5730\u8fdb\u884c\u6570\u636e\u7684\u6807\u51c6\u5316\u3001\u62df\u5408\u548c\u8f6c\u6362\u3002\u5177\u4f53\u6b65\u9aa4\u5305\u62ec\uff1a\u9996\u5148\uff0c\u5bfc\u5165\u6240\u9700\u7684\u5e93\uff1b\u5176\u6b21\uff0c\u52a0\u8f7d\u6570\u636e\u5e76\u8fdb\u884c\u6807\u51c6\u5316\u5904\u7406\uff1b\u7136\u540e\uff0c\u521b\u5efaPCA<\/code>\u5bf9\u8c61\u5e76\u6307\u5b9a\u8981\u4fdd\u7559\u7684\u4e3b\u6210\u5206\u6570\u91cf\uff1b\u6700\u540e\uff0c\u8c03\u7528fit_transform()<\/code>\u65b9\u6cd5\u83b7\u53d6\u964d\u7ef4\u540e\u7684\u6570\u636e\u3002<\/p>\nPCA\u5206\u6790\u540e\u5982\u4f55\u89e3\u91ca\u7ed3\u679c\uff1f<\/strong>
PCA\u7684\u7ed3\u679c\u53ef\u4ee5\u901a\u8fc7\u4e3b\u6210\u5206\u7684\u65b9\u5dee\u89e3\u91ca\u6765\u7406\u89e3\u3002\u6bcf\u4e2a\u4e3b\u6210\u5206\u4ee3\u8868\u539f\u59cb\u6570\u636e\u4e2d\u65b9\u5dee\u7684\u4e00\u4e2a\u65b9\u5411\uff0c\u7528\u6237\u53ef\u4ee5\u67e5\u770b\u6bcf\u4e2a\u4e3b\u6210\u5206\u6240\u5360\u7684\u65b9\u5dee\u6bd4\u4f8b\uff0c\u4ee5\u5224\u65ad\u54ea\u4e9b\u4e3b\u6210\u5206\u662f\u6700\u91cd\u8981\u7684\u3002\u6b64\u5916\uff0c\u901a\u8fc7\u7ed8\u5236\u4e3b\u6210\u5206\u7684\u6563\u70b9\u56fe\uff0c\u53ef\u4ee5\u76f4\u89c2\u5730\u89c2\u5bdf\u6570\u636e\u7684\u5206\u5e03\u548c\u805a\u7c7b\u60c5\u51b5\uff0c\u4ece\u800c\u5e2e\u52a9\u8bc6\u522b\u6570\u636e\u4e2d\u7684\u6f5c\u5728\u6a21\u5f0f\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u5728Python\u4e2d\u8fdb\u884cPCA\uff08\u4e3b\u6210\u5206\u5206\u6790\uff09\uff0c\u53ef\u4ee5\u4f7f\u7528Pandas\u8fdb\u884c\u6570\u636e\u9884\u5904\u7406\u3001\u4f7f\u7528Scikit-learn\u5e93\u6267 […]","protected":false},"author":3,"featured_media":992373,"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\/992366"}],"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=992366"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/992366\/revisions"}],"predecessor-version":[{"id":992374,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/992366\/revisions\/992374"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/992373"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=992366"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=992366"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=992366"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"python](\"https:\/\/cdn-kb.worktile.com\/kb\/wp-content\/uploads\/2024\/04\/25070231\/ca3c35bd-0998-42b7-9c5a-f41fc9fbdaf7.webp\")
<\/p>\n