{
  "id": 392521,
  "title": "Asume same order of landmarks?",
  "url": "/competitions/asl-signs/discussion/392521",
  "author_name": "",
  "post_date": "2023-03-05T17:20:29.204020300Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>The Evaluation states that a video is processed as follows:</p>\n<pre><code>ROWS_PER_FRAME = 543  # number of landmarks per frame\n\ndef load_relevant_data_subset(pq_path):\n    data_columns = ['x', 'y', 'z']\n    data = pd.read_parquet(pq_path, columns=data_columns)\n    n_frames = int(len(data) / ROWS_PER_FRAME)\n    data = data.values.reshape(n_frames, ROWS_PER_FRAME, len(data_columns))\n    return data.astype(np.float32)\n</code></pre>\n<p>Does this mean that we can safely assume that the order of the landmark indexes as well as the types will be mantained at inference time?</p>",
  "messages": [
    {
      "id": "2170096",
      "postDate": "03/05/2023 17:20:29",
      "content": "<p>The Evaluation states that a video is processed as follows:</p>\n<pre><code>ROWS_PER_FRAME = 543  # number of landmarks per frame\n\ndef load_relevant_data_subset(pq_path):\n    data_columns = ['x', 'y', 'z']\n    data = pd.read_parquet(pq_path, columns=data_columns)\n    n_frames = int(len(data) / ROWS_PER_FRAME)\n    data = data.values.reshape(n_frames, ROWS_PER_FRAME, len(data_columns))\n    return data.astype(np.float32)\n</code></pre>\n<p>Does this mean that we can safely assume that the order of the landmark indexes as well as the types will be mantained at inference time?</p>",
      "rawMarkdown": "The Evaluation states that a video is processed as follows:\n```\nROWS_PER_FRAME = 543  # number of landmarks per frame\n\ndef load_relevant_data_subset(pq_path):\n    data_columns = ['x', 'y', 'z']\n    data = pd.read_parquet(pq_path, columns=data_columns)\n    n_frames = int(len(data) / ROWS_PER_FRAME)\n    data = data.values.reshape(n_frames, ROWS_PER_FRAME, len(data_columns))\n    return data.astype(np.float32)\n```\n\nDoes this mean that we can safely assume that the order of the landmark indexes as well as the types will be mantained at inference time?",
      "votes": null
    },
    {
      "id": "2170184",
      "postDate": "03/05/2023 18:37:02",
      "content": "<p>In case anyone is interested, I have just published a Notebook that confirms that the data is arranged the same for all videos, (at least for the training dataset).<br>\n<a href=\"https://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same\" target=\"_blank\">https://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same</a></p>",
      "rawMarkdown": "In case anyone is interested, I have just published a Notebook that confirms that the data is arranged the same for all videos, (at least for the training dataset).\nhttps://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2170184,
      "author_name": "juanumusic",
      "author_url": "",
      "post_date": "03/05/2023 18:37:02",
      "content": "<p>In case anyone is interested, I have just published a Notebook that confirms that the data is arranged the same for all videos, (at least for the training dataset).<br>\n<a href=\"https://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same\" target=\"_blank\">https://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2170096": "The Evaluation states that a video is processed as follows:\n```\nROWS_PER_FRAME = 543  # number of landmarks per frame\n\ndef load_relevant_data_subset(pq_path):\n    data_columns = ['x', 'y', 'z']\n    data = pd.read_parquet(pq_path, columns=data_columns)\n    n_frames = int(len(data) / ROWS_PER_FRAME)\n    data = data.values.reshape(n_frames, ROWS_PER_FRAME, len(data_columns))\n    return data.astype(np.float32)\n```\n\nDoes this mean that we can safely assume that the order of the landmark indexes as well as the types will be mantained at inference time?",
    "2170184": "In case anyone is interested, I have just published a Notebook that confirms that the data is arranged the same for all videos, (at least for the training dataset).\nhttps://www.kaggle.com/juanumusic/validate-all-data-is-arranged-the-same"
  },
  "source": "meta"
}