{
  "id": 66256,
  "title": "What is the encoding of 'class-descriptions.csv' ?",
  "url": "/competitions/inclusive-images-challenge/discussion/66256",
  "author_name": "",
  "post_date": "2018-09-19T17:16:48.361969500Z",
  "votes": null,
  "comment_count": 7,
  "views": 0,
  "content": "<p>Hi, I've attempted to open the file <code>class-descriptions.csv</code> in multiple encodings (UTF-8, ISO 8859-1, etc.) but I haven't been able to get meaningful text of some class descriptions, some examples:</p>\n\n<p>/m/01dnv_,Bağlama\n/m/01csdn,SmörgÃ¥sbord\n/m/01d1dj,Coupé (should this be coupé?)</p>\n\n<p>Hey @organizers, can you let us know what the right encoding is?</p>",
  "messages": [
    {
      "id": "390077",
      "postDate": "09/19/2018 17:16:48",
      "content": "<p>Hi, I've attempted to open the file <code>class-descriptions.csv</code> in multiple encodings (UTF-8, ISO 8859-1, etc.) but I haven't been able to get meaningful text of some class descriptions, some examples:</p>\n\n<p>/m/01dnv_,Bağlama\n/m/01csdn,SmörgÃ¥sbord\n/m/01d1dj,Coupé (should this be coupé?)</p>\n\n<p>Hey @organizers, can you let us know what the right encoding is?</p>",
      "rawMarkdown": "Hi, I've attempted to open the file `class-descriptions.csv` in multiple encodings (UTF-8, ISO 8859-1, etc.) but I haven't been able to get meaningful text of some class descriptions, some examples:\n\n/m/01dnv_,Bağlama\n/m/01csdn,SmörgÃ¥sbord\n/m/01d1dj,Coupé (should this be coupé?)\n\nHey @organizers, can you let us know what the right encoding is?",
      "votes": null
    },
    {
      "id": "390094",
      "postDate": "09/19/2018 17:42:10",
      "content": "<p>I think that either your local copy of the data is corrupted or it's an issue with whatever tool you're using to read the data. I just downloaded that file and was able to read it without problems:</p>\n\n<pre><code>&gt;&gt;&gt; df = pd.read_csv('class-descriptions.csv')\n&gt;&gt;&gt; df.head()\n    label_code          description\n0  /m/0100nhbf     Sprenger's tulip\n1  /m/0104x9kv             Vinegret\n2  /m/0105jzwx            Dabu-dabu\n3  /m/0105ld7g  Pistachio ice cream\n4  /m/0105lxy5                 Woku\n</code></pre>",
      "rawMarkdown": "I think that either your local copy of the data is corrupted or it's an issue with whatever tool you're using to read the data. I just downloaded that file and was able to read it without problems:\n\n    &gt;&gt;&gt; df = pd.read_csv('class-descriptions.csv')\n    &gt;&gt;&gt; df.head()\n        label_code          description\n    0  /m/0100nhbf     Sprenger's tulip\n    1  /m/0104x9kv             Vinegret\n    2  /m/0105jzwx            Dabu-dabu\n    3  /m/0105ld7g  Pistachio ice cream\n    4  /m/0105lxy5                 Woku",
      "votes": null
    },
    {
      "id": "390102",
      "postDate": "09/19/2018 18:00:49",
      "content": "<p>Can you do:</p>\n\n<p>image: <img src=\"http://i.imgur.com/gj7O9h1.png\" alt=\"enter image description here\">\n(btw I give up using code syntax in Kaggle's editor, I am totally unable to do it)</p>",
      "rawMarkdown": "Can you do:\n\nimage: ![enter image description here][1]\n(btw I give up using code syntax in Kaggle's editor, I am totally unable to do it)\n\n\n  [1]: http://i.imgur.com/gj7O9h1.png",
      "votes": null
    },
    {
      "id": "390470",
      "postDate": "09/20/2018 09:07:49",
      "content": "<p>To reiterate: most of the items are OK, but some (e.g. <code>df.loc[993]</code> or <code>df.loc[1017]</code>) contain unreadable characters.</p>",
      "rawMarkdown": "To reiterate: most of the items are OK, but some (e.g. `df.loc[993]` or `df.loc[1017]`) contain unreadable characters.",
      "votes": null
    },
    {
      "id": "392738",
      "postDate": "09/24/2018 08:56:51",
      "content": "<p>@Pallavi can you comment? (specifically about some items as described below)...</p>",
      "rawMarkdown": "Pallavi can you comment? (specifically about some items as described below)...",
      "votes": null
    },
    {
      "id": "393448",
      "postDate": "09/25/2018 11:46:39",
      "content": "<p>I do not see the errors you are seeing   Could you try downloading the file again. </p>",
      "rawMarkdown": "I do not see the errors you are seeing   Could you try downloading the file again.",
      "votes": null
    },
    {
      "id": "393462",
      "postDate": "09/25/2018 12:13:03",
      "content": "<p>Just downloaded again using Kaggle API, same:</p>\n\n<p><img src=\"http://i.imgur.com/OeYlFwt.png\" alt=\"enter image description here\"></p>",
      "rawMarkdown": "Just downloaded again using Kaggle API, same:\n\n![enter image description here][1]\n\n\n  [1]: http://i.imgur.com/OeYlFwt.png",
      "votes": null
    },
    {
      "id": "405206",
      "postDate": "10/17/2018 03:58:35",
      "content": "<p>@Pallavi   @Sohier   my kernel list all none-english description: <a href=\"https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions\">https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions</a></p>",
      "rawMarkdown": "Pallavi   @Sohier   my kernel list all none-english description: https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 390094,
      "author_name": "sohier",
      "author_url": "",
      "post_date": "09/19/2018 17:42:10",
      "content": "<p>I think that either your local copy of the data is corrupted or it's an issue with whatever tool you're using to read the data. I just downloaded that file and was able to read it without problems:</p>\n\n<pre><code>&gt;&gt;&gt; df = pd.read_csv('class-descriptions.csv')\n&gt;&gt;&gt; df.head()\n    label_code          description\n0  /m/0100nhbf     Sprenger's tulip\n1  /m/0104x9kv             Vinegret\n2  /m/0105jzwx            Dabu-dabu\n3  /m/0105ld7g  Pistachio ice cream\n4  /m/0105lxy5                 Woku\n</code></pre>",
      "votes": null,
      "replies": [
        {
          "id": 390102,
          "author_name": "antorsae",
          "author_url": "",
          "post_date": "09/19/2018 18:00:49",
          "content": "<p>Can you do:</p>\n\n<p>image: <img src=\"http://i.imgur.com/gj7O9h1.png\" alt=\"enter image description here\">\n(btw I give up using code syntax in Kaggle's editor, I am totally unable to do it)</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 390470,
          "author_name": "antorsae",
          "author_url": "",
          "post_date": "09/20/2018 09:07:49",
          "content": "<p>To reiterate: most of the items are OK, but some (e.g. <code>df.loc[993]</code> or <code>df.loc[1017]</code>) contain unreadable characters.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 392738,
      "author_name": "antorsae",
      "author_url": "",
      "post_date": "09/24/2018 08:56:51",
      "content": "<p>@Pallavi can you comment? (specifically about some items as described below)...</p>",
      "votes": null,
      "replies": [
        {
          "id": 393448,
          "author_name": "pbaljeka",
          "author_url": "",
          "post_date": "09/25/2018 11:46:39",
          "content": "<p>I do not see the errors you are seeing   Could you try downloading the file again. </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 393462,
          "author_name": "antorsae",
          "author_url": "",
          "post_date": "09/25/2018 12:13:03",
          "content": "<p>Just downloaded again using Kaggle API, same:</p>\n\n<p><img src=\"http://i.imgur.com/OeYlFwt.png\" alt=\"enter image description here\"></p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 405206,
      "author_name": "yyqing",
      "author_url": "",
      "post_date": "10/17/2018 03:58:35",
      "content": "<p>@Pallavi   @Sohier   my kernel list all none-english description: <a href=\"https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions\">https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "390077": "Hi, I've attempted to open the file `class-descriptions.csv` in multiple encodings (UTF-8, ISO 8859-1, etc.) but I haven't been able to get meaningful text of some class descriptions, some examples:\n\n/m/01dnv_,Bağlama\n/m/01csdn,SmörgÃ¥sbord\n/m/01d1dj,Coupé (should this be coupé?)\n\nHey @organizers, can you let us know what the right encoding is?",
    "390094": "I think that either your local copy of the data is corrupted or it's an issue with whatever tool you're using to read the data. I just downloaded that file and was able to read it without problems:\n\n    &gt;&gt;&gt; df = pd.read_csv('class-descriptions.csv')\n    &gt;&gt;&gt; df.head()\n        label_code          description\n    0  /m/0100nhbf     Sprenger's tulip\n    1  /m/0104x9kv             Vinegret\n    2  /m/0105jzwx            Dabu-dabu\n    3  /m/0105ld7g  Pistachio ice cream\n    4  /m/0105lxy5                 Woku",
    "390102": "Can you do:\n\nimage: ![enter image description here][1]\n(btw I give up using code syntax in Kaggle's editor, I am totally unable to do it)\n\n\n  [1]: http://i.imgur.com/gj7O9h1.png",
    "390470": "To reiterate: most of the items are OK, but some (e.g. `df.loc[993]` or `df.loc[1017]`) contain unreadable characters.",
    "392738": "Pallavi can you comment? (specifically about some items as described below)...",
    "393448": "I do not see the errors you are seeing   Could you try downloading the file again.",
    "393462": "Just downloaded again using Kaggle API, same:\n\n![enter image description here][1]\n\n\n  [1]: http://i.imgur.com/OeYlFwt.png",
    "405206": "Pallavi   @Sohier   my kernel list all none-english description: https://www.kaggle.com/yyqing/strange-chars-in-class-descriptions"
  },
  "source": "meta"
}