{
  "id": 26959,
  "title": "Docker image python 2",
  "url": "/competitions/dstl-satellite-imagery-feature-detection/discussion/26959",
  "author_name": "",
  "post_date": "2016-12-26T22:15:12.770Z",
  "votes": 1,
  "comment_count": 3,
  "views": 180,
  "content": "<p>I'm new to Kaggle and Docker, so wanted to check something. After executing the command.sh the Jupyter notebook that's generated seems to be using Python 2. Is this because Kaggle have set it that way or is it my own settings not being right?</p>\n\n<p>I'm trying to maintain all my code in Python 3 so would prefer to use that.</p>",
  "messages": [
    {
      "id": "152521",
      "postDate": "12/26/2016 22:15:12",
      "content": "<p>I'm new to Kaggle and Docker, so wanted to check something. After executing the command.sh the Jupyter notebook that's generated seems to be using Python 2. Is this because Kaggle have set it that way or is it my own settings not being right?</p>\n\n<p>I'm trying to maintain all my code in Python 3 so would prefer to use that.</p>",
      "rawMarkdown": "I'm new to Kaggle and Docker, so wanted to check something. After executing the command.sh the Jupyter notebook that's generated seems to be using Python 2. Is this because Kaggle have set it that way or is it my own settings not being right?\r\n\r\nI'm trying to maintain all my code in Python 3 so would prefer to use that.",
      "votes": null
    },
    {
      "id": "152549",
      "postDate": "12/27/2016 03:57:56",
      "content": "<p>I've been able to use Python 3.5 with the kaggle/python image.  After pulling the image, I use code like this in my shell:</p>\n\n<pre><code>docker run --rm -it -p 8888:8888  -v d:/Kaggles:/d  kaggle/python\njupyter notebook --ip=0.0.0.0 --no-browser\n</code></pre>\n\n<p>Then I go to the browser and hit <a href=\"https://localhost:8888/\">https://localhost:8888</a>, and when I open a new kernel it's with Python 3.5/ I don't recall doing anything special when pulling the image or setting up Docker.</p>",
      "rawMarkdown": "I've been able to use Python 3.5 with the kaggle/python image.  After pulling the image, I use code like this in my shell:\r\n\r\n    docker run --rm -it -p 8888:8888  -v d:/Kaggles:/d  kaggle/python\r\n    jupyter notebook --ip=0.0.0.0 --no-browser\r\n\r\nThen I go to the browser and hit https://localhost:8888, and when I open a new kernel it's with Python 3.5/ I don't recall doing anything special when pulling the image or setting up Docker.",
      "votes": null
    },
    {
      "id": "152642",
      "postDate": "12/27/2016 17:44:19",
      "content": "<p>Hmm strange. I just followed the Kaggle image pull instructions, followed by downloading this and running command.sh <a href=\"https://github.com/wendykan/docker-python-GDAL\">https://github.com/wendykan/docker-python-GDAL</a></p>\n\n<p>Not sure why my notebook only gives me Python 2</p>\n\n<p>Update: Ok I think it's because I mistakenly assumed the Docker image provided for the DSTL challenge somehow sits within the Kaggle image. Turns out its an entirely separate image.</p>\n\n<p>I guess I'll remove it and stick to the Kaggle one.</p>",
      "rawMarkdown": "Hmm strange. I just followed the Kaggle image pull instructions, followed by downloading this and running command.sh https://github.com/wendykan/docker-python-GDAL\r\n\r\nNot sure why my notebook only gives me Python 2\r\n\r\nUpdate: Ok I think it's because I mistakenly assumed the Docker image provided for the DSTL challenge somehow sits within the Kaggle image. Turns out its an entirely separate image.\r\n\r\nI guess I'll remove it and stick to the Kaggle one.",
      "votes": null
    },
    {
      "id": "152645",
      "postDate": "12/27/2016 17:56:50",
      "content": "<p>It looks like you're building your own image called python-gdal and then running the container based on that. You might try just \"docker pull kaggle/python\" which will get the latest kaggle image from Docker Hub. Then run the commands I listed above, adjusted for your OS and setup. </p>",
      "rawMarkdown": "It looks like you're building your own image called python-gdal and then running the container based on that. You might try just \"docker pull kaggle/python\" which will get the latest kaggle image from Docker Hub. Then run the commands I listed above, adjusted for your OS and setup.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 152549,
      "author_name": "jpmiller",
      "author_url": "",
      "post_date": "12/27/2016 03:57:56",
      "content": "<p>I've been able to use Python 3.5 with the kaggle/python image.  After pulling the image, I use code like this in my shell:</p>\n\n<pre><code>docker run --rm -it -p 8888:8888  -v d:/Kaggles:/d  kaggle/python\njupyter notebook --ip=0.0.0.0 --no-browser\n</code></pre>\n\n<p>Then I go to the browser and hit <a href=\"https://localhost:8888/\">https://localhost:8888</a>, and when I open a new kernel it's with Python 3.5/ I don't recall doing anything special when pulling the image or setting up Docker.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 152642,
      "author_name": "badrul",
      "author_url": "",
      "post_date": "12/27/2016 17:44:19",
      "content": "<p>Hmm strange. I just followed the Kaggle image pull instructions, followed by downloading this and running command.sh <a href=\"https://github.com/wendykan/docker-python-GDAL\">https://github.com/wendykan/docker-python-GDAL</a></p>\n\n<p>Not sure why my notebook only gives me Python 2</p>\n\n<p>Update: Ok I think it's because I mistakenly assumed the Docker image provided for the DSTL challenge somehow sits within the Kaggle image. Turns out its an entirely separate image.</p>\n\n<p>I guess I'll remove it and stick to the Kaggle one.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 152645,
      "author_name": "jpmiller",
      "author_url": "",
      "post_date": "12/27/2016 17:56:50",
      "content": "<p>It looks like you're building your own image called python-gdal and then running the container based on that. You might try just \"docker pull kaggle/python\" which will get the latest kaggle image from Docker Hub. Then run the commands I listed above, adjusted for your OS and setup. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "152521": "I'm new to Kaggle and Docker, so wanted to check something. After executing the command.sh the Jupyter notebook that's generated seems to be using Python 2. Is this because Kaggle have set it that way or is it my own settings not being right?\r\n\r\nI'm trying to maintain all my code in Python 3 so would prefer to use that.",
    "152549": "I've been able to use Python 3.5 with the kaggle/python image.  After pulling the image, I use code like this in my shell:\r\n\r\n    docker run --rm -it -p 8888:8888  -v d:/Kaggles:/d  kaggle/python\r\n    jupyter notebook --ip=0.0.0.0 --no-browser\r\n\r\nThen I go to the browser and hit https://localhost:8888, and when I open a new kernel it's with Python 3.5/ I don't recall doing anything special when pulling the image or setting up Docker.",
    "152642": "Hmm strange. I just followed the Kaggle image pull instructions, followed by downloading this and running command.sh https://github.com/wendykan/docker-python-GDAL\r\n\r\nNot sure why my notebook only gives me Python 2\r\n\r\nUpdate: Ok I think it's because I mistakenly assumed the Docker image provided for the DSTL challenge somehow sits within the Kaggle image. Turns out its an entirely separate image.\r\n\r\nI guess I'll remove it and stick to the Kaggle one.",
    "152645": "It looks like you're building your own image called python-gdal and then running the container based on that. You might try just \"docker pull kaggle/python\" which will get the latest kaggle image from Docker Hub. Then run the commands I listed above, adjusted for your OS and setup."
  },
  "source": "meta"
}