{
  "id": 626888,
  "title": "train.csv in the root directory?",
  "url": "/competitions/physionet-ecg-image-digitization/discussion/626888",
  "author_name": "",
  "post_date": "2025-11-16T15:09:45.867549100Z",
  "votes": -3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Folks I see a train.csv in the root directory of physionet-ecg-image-digitization.zip. Since there is a .csv in each of the 768 sub-directories under train, which images does the data in train.csv correspond to? It appears to have time-series data.</p>\n<p>$ ls\nsample_submission.parquet  test  test.csv  train  train.csv\n$ head train.csv\nid,fs,sig_len\n7663343,500,5000\n10140238,1000,10000\n11842146,1000,10000</p>",
  "messages": [
    {
      "id": "3331987",
      "postDate": "11/16/2025 15:09:45",
      "content": "<p>Folks I see a train.csv in the root directory of physionet-ecg-image-digitization.zip. Since there is a .csv in each of the 768 sub-directories under train, which images does the data in train.csv correspond to? It appears to have time-series data.</p>\n<p>$ ls\nsample_submission.parquet  test  test.csv  train  train.csv\n$ head train.csv\nid,fs,sig_len\n7663343,500,5000\n10140238,1000,10000\n11842146,1000,10000</p>",
      "rawMarkdown": "Folks I see a train.csv in the root directory of physionet-ecg-image-digitization.zip. Since there is a .csv in each of the 768 sub-directories under train, which images does the data in train.csv correspond to? It appears to have time-series data.\n\n$ ls\nsample_submission.parquet  test  test.csv  train  train.csv\n$ head train.csv\nid,fs,sig_len\n7663343,500,5000\n10140238,1000,10000\n11842146,1000,10000",
      "votes": null
    },
    {
      "id": "3332293",
      "postDate": "11/16/2025 16:56:13",
      "content": "<p>The train.csv in the root folder is not the actual ECG signal.\nIt’s just a metadata file, where each row corresponds to one record/folder inside the train/ directory.</p>\n<p>The real time-series signals are stored in the per-record CSV files, for example:\ntrain/12345/12345.csv</p>\n<p>Those files contain the full 12-lead waveform.</p>\n<p>So the root train.csv tells you which record IDs exist and provides some basic info, while the actual signals are in the train//.csv files. Use the id column in the root file to match them.</p>",
      "rawMarkdown": "The train.csv in the root folder is not the actual ECG signal.\nIt’s just a metadata file, where each row corresponds to one record/folder inside the train/ directory.\n\nThe real time-series signals are stored in the per-record CSV files, for example:\ntrain/12345/12345.csv\n\nThose files contain the full 12-lead waveform.\n\nSo the root train.csv tells you which record IDs exist and provides some basic info, while the actual signals are in the train/<id>/<id>.csv files. Use the id column in the root file to match them.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3332293,
      "author_name": "trishalaas",
      "author_url": "",
      "post_date": "11/16/2025 16:56:13",
      "content": "<p>The train.csv in the root folder is not the actual ECG signal.\nIt’s just a metadata file, where each row corresponds to one record/folder inside the train/ directory.</p>\n<p>The real time-series signals are stored in the per-record CSV files, for example:\ntrain/12345/12345.csv</p>\n<p>Those files contain the full 12-lead waveform.</p>\n<p>So the root train.csv tells you which record IDs exist and provides some basic info, while the actual signals are in the train//.csv files. Use the id column in the root file to match them.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3331987": "Folks I see a train.csv in the root directory of physionet-ecg-image-digitization.zip. Since there is a .csv in each of the 768 sub-directories under train, which images does the data in train.csv correspond to? It appears to have time-series data.\n\n$ ls\nsample_submission.parquet  test  test.csv  train  train.csv\n$ head train.csv\nid,fs,sig_len\n7663343,500,5000\n10140238,1000,10000\n11842146,1000,10000",
    "3332293": "The train.csv in the root folder is not the actual ECG signal.\nIt’s just a metadata file, where each row corresponds to one record/folder inside the train/ directory.\n\nThe real time-series signals are stored in the per-record CSV files, for example:\ntrain/12345/12345.csv\n\nThose files contain the full 12-lead waveform.\n\nSo the root train.csv tells you which record IDs exist and provides some basic info, while the actual signals are in the train/<id>/<id>.csv files. Use the id column in the root file to match them."
  },
  "source": "meta"
}