{"id":1145937,"date":"2025-01-08T23:12:33","date_gmt":"2025-01-08T15:12:33","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1145937.html"},"modified":"2025-01-08T23:12:36","modified_gmt":"2025-01-08T15:12:36","slug":"python%e5%a6%82%e4%bd%95%e5%b0%86%e7%9f%a9%e9%98%b5%e5%ad%98%e5%82%a8%e4%b8%bah5%e6%96%87%e4%bb%b6","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/1145937.html","title":{"rendered":"python\u5982\u4f55\u5c06\u77e9\u9635\u5b58\u50a8\u4e3ah5\u6587\u4ef6"},"content":{"rendered":"

\"python\u5982\u4f55\u5c06\u77e9\u9635\u5b58\u50a8\u4e3ah5\u6587\u4ef6\"<\/p>\n

Python\u5c06\u77e9\u9635\u5b58\u50a8\u4e3aH5\u6587\u4ef6\u7684\u65b9\u6cd5<\/strong>\uff1a\u4f7f\u7528HDF5\u5e93\u3001\u4f7f\u7528h5py\u5e93\u3001\u5229\u7528Pandas\u5e93\u3001\u4f7f\u7528NumPy\u5e93\u3002\u63a8\u8350\u4f7f\u7528h5py\u5e93\uff0c\u56e0\u4e3a\u5b83\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7b80\u5355\u4e14\u9ad8\u6548\u7684\u63a5\u53e3\u6765\u5904\u7406HDF5\u6587\u4ef6<\/strong>\u3002<\/p>\n<\/p>\n

\u4e00\u3001HDF5\u6587\u4ef6\u7b80\u4ecb<\/h3>\n<\/p>\n

HDF5\uff08Hierarchical Data Format version 5\uff09\u662f\u4e00\u79cd\u7528\u4e8e\u5b58\u50a8\u548c\u7ec4\u7ec7\u5927\u89c4\u6a21\u6570\u636e\u7684\u6587\u4ef6\u683c\u5f0f\u3002HDF5\u6587\u4ef6\u53ef\u4ee5\u5305\u542b\u4efb\u610f\u7c7b\u578b\u7684\u6570\u636e\u96c6\uff0c\u5e76\u4e14\u652f\u6301\u591a\u79cd\u6570\u636e\u7c7b\u578b\u548c\u590d\u6742\u7684\u6570\u636e\u7ed3\u6784\u3002\u5176\u4e3b\u8981\u4f18\u52bf\u5305\u62ec\u6570\u636e\u538b\u7f29\u3001\u8de8\u5e73\u53f0\u517c\u5bb9\u6027\u548c\u5f3a\u5927\u7684\u5143\u6570\u636e\u652f\u6301\u3002<\/p>\n<\/p>\n

\u4e8c\u3001\u5b89\u88c5h5py\u5e93<\/h3>\n<\/p>\n

\u5728\u5f00\u59cb\u4e4b\u524d\uff0c\u60a8\u9700\u8981\u786e\u4fdd\u5df2\u5b89\u88c5h5py\u5e93\u3002\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u547d\u4ee4\u8fdb\u884c\u5b89\u88c5\uff1a<\/p>\n<\/p>\n

pip install h5py<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001\u521b\u5efa\u5e76\u5b58\u50a8\u77e9\u9635<\/h3>\n<\/p>\n
1. \u521b\u5efa\u77e9\u9635<\/h4>\n<\/p>\n
\u4f7f\u7528NumPy\u5e93\u521b\u5efa\u4e00\u4e2a\u77e9\u9635\u3002NumPy\u662fPython\u4e2d\u5904\u7406\u6570\u7ec4\u548c\u77e9\u9635\u7684\u6807\u51c6\u5e93\u3002<\/p>\n<\/p>\n
import numpy as np<\/p>\n
\u521b\u5efa\u4e00\u4e2a\u968f\u673a\u77e9\u9635<\/strong><\/h2>\n
matrix = np.random.rand(5, 5)<\/p>\n
<\/code><\/pre>\n<\/p>\n
2. \u4f7f\u7528h5py\u5e93\u5b58\u50a8\u77e9\u9635<\/h4>\n<\/p>\n
h5py\u5e93\u63d0\u4f9b\u4e86\u4e00\u79cd\u7b80\u5355\u7684\u65b9\u6cd5\u6765\u5c06\u77e9\u9635\u5b58\u50a8\u4e3aH5\u6587\u4ef6\u3002<\/p>\n<\/p>\n
import h5py<\/p>\n
\u521b\u5efa\u4e00\u4e2aHDF5\u6587\u4ef6<\/strong><\/h2>\n
with h5py.File('matrix.h5', 'w') as f:<\/p>\n
    # \u521b\u5efa\u4e00\u4e2a\u6570\u636e\u96c6\u5e76\u5b58\u50a8\u77e9\u9635<\/p>\n
    f.create_dataset('dataset_1', data=matrix)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u56db\u3001\u8bfb\u53d6H5\u6587\u4ef6\u4e2d\u7684\u77e9\u9635<\/h3>\n<\/p>\n
\u540c\u6837\uff0c\u4f7f\u7528h5py\u5e93\u8bfb\u53d6\u5b58\u50a8\u5728H5\u6587\u4ef6\u4e2d\u7684\u77e9\u9635\u3002<\/p>\n<\/p>\n
with h5py.File('matrix.h5', 'r') as f:<\/p>\n
    # \u8bfb\u53d6\u6570\u636e\u96c6<\/p>\n
    data = f['dataset_1'][:]<\/p>\n
    print(data)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e94\u3001\u4f7f\u7528Pandas\u5b58\u50a8\u548c\u8bfb\u53d6\u77e9\u9635<\/h3>\n<\/p>\n
\u867d\u7136h5py\u662f\u5904\u7406HDF5\u6587\u4ef6\u7684\u4e3b\u8981\u5e93\uff0c\u4f46Pandas\u5e93\u4e5f\u63d0\u4f9b\u4e86\u5bf9HDF5\u6587\u4ef6\u7684\u652f\u6301\uff0c\u7279\u522b\u662f\u5728\u5904\u7406DataFrame\u65f6\u3002<\/p>\n<\/p>\n
1. \u4f7f\u7528Pandas\u5b58\u50a8\u77e9\u9635<\/h4>\n<\/p>\n
import pandas as pd<\/p>\n
\u5c06\u77e9\u9635\u8f6c\u6362\u4e3aDataFrame<\/strong><\/h2>\n
df = pd.DataFrame(matrix)<\/p>\n
\u5c06DataFrame\u5b58\u50a8\u4e3aHDF5\u6587\u4ef6<\/strong><\/h2>\n
df.to_hdf('matrix_pandas.h5', key='df', mode='w')<\/p>\n
<\/code><\/pre>\n<\/p>\n
2. \u4f7f\u7528Pandas\u8bfb\u53d6\u77e9\u9635<\/h4>\n<\/p>\n
# \u8bfb\u53d6HDF5\u6587\u4ef6\u4e2d\u7684DataFrame<\/p>\n
df_read = pd.read_hdf('matrix_pandas.h5', 'df')<\/p>\n
print(df_read)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u516d\u3001HDF5\u6587\u4ef6\u7684\u9ad8\u7ea7\u64cd\u4f5c<\/h3>\n<\/p>\n
1. \u6570\u636e\u538b\u7f29<\/h4>\n<\/p>\n
HDF5\u6587\u4ef6\u652f\u6301\u6570\u636e\u538b\u7f29\uff0c\u53ef\u4ee5\u663e\u8457\u51cf\u5c11\u6587\u4ef6\u5927\u5c0f\u3002<\/p>\n<\/p>\n
with h5py.File('matrix_compressed.h5', 'w') as f:<\/p>\n
    f.create_dataset('dataset_1', data=matrix, compression='gzip')<\/p>\n
<\/code><\/pre>\n<\/p>\n
2. \u6dfb\u52a0\u5143\u6570\u636e<\/h4>\n<\/p>\n
\u53ef\u4ee5\u5411HDF5\u6587\u4ef6\u6216\u6570\u636e\u96c6\u6dfb\u52a0\u5143\u6570\u636e\uff0c\u4ee5\u4fbf\u66f4\u597d\u5730\u63cf\u8ff0\u6570\u636e\u3002<\/p>\n<\/p>\n
with h5py.File('matrix_metadata.h5', 'w') as f:<\/p>\n
    dset = f.create_dataset('dataset_1', data=matrix)<\/p>\n
    dset.attrs['description'] = 'This is a 5x5 random matrix'<\/p>\n
    dset.attrs['author'] = 'Your Name'<\/p>\n
<\/code><\/pre>\n<\/p>\n
3. \u521b\u5efa\u5206\u5c42\u7ed3\u6784<\/h4>\n<\/p>\n
HDF5\u6587\u4ef6\u5141\u8bb8\u521b\u5efa\u5206\u5c42\u7ed3\u6784\uff0c\u4ee5\u4fbf\u66f4\u597d\u5730\u7ec4\u7ec7\u6570\u636e\u3002<\/p>\n<\/p>\n
with h5py.File('matrix_hierarchy.h5', 'w') as f:<\/p>\n
    grp = f.create_group('group1')<\/p>\n
    grp.create_dataset('dataset_1', data=matrix)<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e03\u3001\u5b9e\u8df5\u4e2d\u7684\u5e94\u7528\u573a\u666f<\/h3>\n<\/p>\n
1. \u79d1\u5b66\u8ba1\u7b97<\/h4>\n<\/p>\n
\u5728\u79d1\u5b66\u8ba1\u7b97\u4e2d\uff0c\u7ecf\u5e38\u9700\u8981\u5904\u7406\u5927\u89c4\u6a21\u7684\u77e9\u9635\u6570\u636e\u3002HDF5\u6587\u4ef6\u7531\u4e8e\u5176\u9ad8\u6548\u7684\u6570\u636e\u5b58\u50a8\u548c\u8bfb\u53d6\u6027\u80fd\uff0c\u6210\u4e3a\u5b58\u50a8\u79d1\u5b66\u6570\u636e\u7684\u9996\u9009\u3002<\/p>\n<\/p>\n
2. \u673a\u5668\u5b66\u4e60<\/a><\/h4>\n<\/p>\n
\u5728\u673a\u5668\u5b66\u4e60\u4e2d\uff0c\u5c24\u5176\u662f\u6df1\u5ea6\u5b66\u4e60\u4e2d\uff0c\u9700\u8981\u5904\u7406\u5927\u91cf\u7684\u8bad\u7ec3\u6570\u636e\u548c\u6a21\u578b\u53c2\u6570\u3002HDF5\u6587\u4ef6\u53ef\u4ee5\u6709\u6548\u5730\u5b58\u50a8\u8fd9\u4e9b\u6570\u636e\uff0c\u5e76\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u5feb\u901f\u8bfb\u53d6\u3002<\/p>\n<\/p>\n
3. \u6570\u636e\u5206\u6790<\/h4>\n<\/p>\n
\u5728\u6570\u636e\u5206\u6790\u4e2d\uff0c\u7ecf\u5e38\u9700\u8981\u5904\u7406\u5927\u89c4\u6a21\u7684\u7ed3\u6784\u5316\u6570\u636e\u3002HDF5\u6587\u4ef6\u652f\u6301\u590d\u6742\u7684\u6570\u636e\u7ed3\u6784\u548c\u5143\u6570\u636e\uff0c\u53ef\u4ee5\u5e2e\u52a9\u6570\u636e\u5206\u6790\u5e08\u66f4\u597d\u5730\u7ec4\u7ec7\u548c\u7ba1\u7406\u6570\u636e\u3002<\/p>\n<\/p>\n
\u516b\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
\u5728Python\u4e2d\uff0c\u5c06\u77e9\u9635\u5b58\u50a8\u4e3aH5\u6587\u4ef6\u662f\u4e00\u79cd\u9ad8\u6548\u4e14\u7075\u6d3b\u7684\u65b9\u6cd5\u3002\u901a\u8fc7\u4f7f\u7528h5py\u5e93\uff0c\u60a8\u53ef\u4ee5\u8f7b\u677e\u5730\u521b\u5efa\u3001\u5b58\u50a8\u548c\u8bfb\u53d6HDF5\u6587\u4ef6<\/strong>\u3002\u6b64\u5916\uff0cPandas\u5e93\u4e5f\u63d0\u4f9b\u4e86\u5bf9HDF5\u6587\u4ef6\u7684\u652f\u6301\uff0c\u7279\u522b\u9002\u7528\u4e8e\u5904\u7406DataFrame\u3002\u65e0\u8bba\u662f\u5728\u79d1\u5b66\u8ba1\u7b97\u3001\u673a\u5668\u5b66\u4e60\u8fd8\u662f\u6570\u636e\u5206\u6790\u4e2d\uff0cHDF5\u6587\u4ef6\u90fd\u53ef\u4ee5\u5e2e\u52a9\u60a8\u66f4\u597d\u5730\u7ba1\u7406\u548c\u5904\u7406\u5927\u89c4\u6a21\u6570\u636e\u3002<\/p>\n<\/p>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
 \u5982\u4f55\u5728Python\u4e2d\u521b\u5efa\u548c\u4fdd\u5b58HDF5\u683c\u5f0f\u7684\u77e9\u9635\uff1f<\/strong>
\u5728Python\u4e2d\uff0c\u53ef\u4ee5\u4f7f\u7528h5py\u5e93\u6765\u521b\u5efa\u548c\u4fdd\u5b58HDF5\u683c\u5f0f\u7684\u77e9\u9635\u3002\u9996\u5148\uff0c\u786e\u4fdd\u5df2\u5b89\u88c5h5py\u5e93\uff0c\u53ef\u4ee5\u901a\u8fc7pip install h5py\u8fdb\u884c\u5b89\u88c5\u3002\u63a5\u4e0b\u6765\uff0c\u521b\u5efa\u4e00\u4e2anumpy\u6570\u7ec4\u8868\u793a\u77e9\u9635\uff0c\u5e76\u4f7f\u7528h5py.File()\u51fd\u6570\u521b\u5efa\u4e00\u4e2aHDF5\u6587\u4ef6\uff0c\u4f7f\u7528.create_dataset()\u65b9\u6cd5\u5c06\u77e9\u9635\u4fdd\u5b58\u5230\u6587\u4ef6\u4e2d\u3002\u793a\u4f8b\u4ee3\u7801\u5982\u4e0b\uff1a<\/p>\n
import numpy as np\nimport h5py\n\n# \u521b\u5efa\u4e00\u4e2a\u793a\u4f8b\u77e9\u9635\nmatrix = np.random.rand(100, 100)\n\n# \u521b\u5efaHDF5\u6587\u4ef6\u5e76\u4fdd\u5b58\u77e9\u9635\nwith h5py.File('matrix.h5', 'w') as hf:\n    hf.create_dataset('my_matrix', data=matrix)\n<\/code><\/pre>\n
HDF5\u6587\u4ef6\u7684\u4f18\u70b9\u662f\u4ec0\u4e48\uff1f<\/strong>
HDF5\u6587\u4ef6\u683c\u5f0f\u5177\u6709\u8bb8\u591a\u4f18\u70b9\uff0c\u4f7f\u5176\u6210\u4e3a\u5b58\u50a8\u5927\u89c4\u6a21\u6570\u636e\u7684\u7406\u60f3\u9009\u62e9\u3002\u9996\u5148\uff0c\u5b83\u652f\u6301\u5b58\u50a8\u5927\u91cf\u6570\u636e\u800c\u4e0d\u53d7\u5185\u5b58\u9650\u5236\uff0c\u9002\u5408\u5904\u7406\u5927\u77e9\u9635\u3002\u5176\u6b21\uff0cHDF5\u6587\u4ef6\u53ef\u4ee5\u8f7b\u677e\u5730\u8fdb\u884c\u6570\u636e\u538b\u7f29\uff0c\u51cf\u5c11\u5b58\u50a8\u7a7a\u95f4\u3002\u6b64\u5916\uff0cHDF5\u652f\u6301\u591a\u79cd\u6570\u636e\u7c7b\u578b\u548c\u9ad8\u6548\u7684\u6570\u636e\u8bbf\u95ee\u65b9\u5f0f\uff0c\u9002\u5408\u79d1\u5b66\u8ba1\u7b97\u548c\u6570\u636e\u5206\u6790\u9886\u57df\u3002<\/p>\n
\u5982\u4f55\u8bfb\u53d6HDF5\u6587\u4ef6\u4e2d\u7684\u77e9\u9635\u6570\u636e\uff1f<\/strong>
\u8bfb\u53d6HDF5\u6587\u4ef6\u4e2d\u7684\u77e9\u9635\u6570\u636e\u540c\u6837\u7b80\u5355\u3002\u4f7f\u7528h5py\u5e93\u6253\u5f00\u6587\u4ef6\uff0c\u5e76\u901a\u8fc7\u6587\u4ef6\u5bf9\u8c61\u8bbf\u95ee\u5df2\u4fdd\u5b58\u7684\u6570\u636e\u96c6\u3002\u4ee5\u4e0b\u793a\u4f8b\u5c55\u793a\u4e86\u5982\u4f55\u8bfb\u53d6\u521a\u624d\u4fdd\u5b58\u7684\u77e9\u9635\uff1a<\/p>\n
with h5py.File('matrix.h5', 'r') as hf:\n    loaded_matrix = hf['my_matrix'][:]\n<\/code><\/pre>\n
\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\uff0c\u60a8\u53ef\u4ee5\u8f7b\u677e\u5730\u5c06HDF5\u6587\u4ef6\u4e2d\u7684\u77e9\u9635\u52a0\u8f7d\u56dePython\u4e2d\u8fdb\u884c\u540e\u7eed\u5904\u7406\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"Python\u5c06\u77e9\u9635\u5b58\u50a8\u4e3aH5\u6587\u4ef6\u7684\u65b9\u6cd5\uff1a\u4f7f\u7528HDF5\u5e93\u3001\u4f7f\u7528h5py\u5e93\u3001\u5229\u7528Pandas\u5e93\u3001\u4f7f\u7528NumPy\u5e93\u3002 […]","protected":false},"author":3,"featured_media":1145949,"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\/1145937"}],"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=1145937"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1145937\/revisions"}],"predecessor-version":[{"id":1145952,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1145937\/revisions\/1145952"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/1145949"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=1145937"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=1145937"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=1145937"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}