{
  "id": 7847,
  "title": "purpose of network position file",
  "url": "/competitions/connectomics/discussion/7847",
  "author_name": "",
  "post_date": "2014-04-24T05:28:48.287Z",
  "votes": null,
  "comment_count": 3,
  "views": 1081,
  "content": "<p>Hi,</p>\n<p>Can anyone explain the purpose of network position file &quot;P&quot; in the data set? Is there any local neighborhood structure between the neurons, and does this data file &quot;P&quot; provides this relationship?&nbsp;</p>\n<p>- Karthik.</p>",
  "messages": [
    {
      "id": "42902",
      "postDate": "04/24/2014 05:28:48",
      "content": "<p>Hi,</p>\n<p>Can anyone explain the purpose of network position file &quot;P&quot; in the data set? Is there any local neighborhood structure between the neurons, and does this data file &quot;P&quot; provides this relationship?&nbsp;</p>\n<p>- Karthik.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "42954",
      "postDate": "04/24/2014 17:05:16",
      "content": "<p>Yes, in the time series data nearby nodes are somewhat correlated&nbsp;to simulate how fluorescence is observed in experiments. &nbsp;Note that P(i-&gt;j) does not depend on distance, however.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43010",
      "postDate": "04/25/2014 15:56:52",
      "content": "<p>But what is then the mapping between the coordinates of the neuron provided in the position file and the neuron indexed by (n1,n2) in the network file?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "43278",
      "postDate": "04/29/2014 08:51:10",
      "content": "<p>The positions file has 2 columns (X and Y coordinates). Each row is a neuron from 1 to N.</p>\n<p>The network file has 3 columns (I,J,W). Each entry denotes a connection from I to J with weight W (the weight only determines if the connection is excitatory +1 or inhibitory -1, although you can disregard the inhibitory ones, since they are blocked in the simulations).</p>\n\n<p>Let's say the positions file starts with:</p>\n<p>0.3 0.5</p>\n<p>0.1 0.7</p>\n<p>And in the network file you see:</p>\n<p>1 2 1</p>\n<p>That would mean that there is a connection from the &nbsp;neuron at position (0.3,0.5)&nbsp;to the neuron at position (0.1, 0.7).</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 42954,
      "author_name": "george04",
      "author_url": "",
      "post_date": "04/24/2014 17:05:16",
      "content": "<p>Yes, in the time series data nearby nodes are somewhat correlated&nbsp;to simulate how fluorescence is observed in experiments. &nbsp;Note that P(i-&gt;j) does not depend on distance, however.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43010,
      "author_name": "",
      "author_url": "",
      "post_date": "04/25/2014 15:56:52",
      "content": "<p>But what is then the mapping between the coordinates of the neuron provided in the position file and the neuron indexed by (n1,n2) in the network file?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 43278,
      "author_name": "orlandi",
      "author_url": "",
      "post_date": "04/29/2014 08:51:10",
      "content": "<p>The positions file has 2 columns (X and Y coordinates). Each row is a neuron from 1 to N.</p>\n<p>The network file has 3 columns (I,J,W). Each entry denotes a connection from I to J with weight W (the weight only determines if the connection is excitatory +1 or inhibitory -1, although you can disregard the inhibitory ones, since they are blocked in the simulations).</p>\n\n<p>Let's say the positions file starts with:</p>\n<p>0.3 0.5</p>\n<p>0.1 0.7</p>\n<p>And in the network file you see:</p>\n<p>1 2 1</p>\n<p>That would mean that there is a connection from the &nbsp;neuron at position (0.3,0.5)&nbsp;to the neuron at position (0.1, 0.7).</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "42902": "",
    "42954": "",
    "43010": "",
    "43278": ""
  },
  "source": "meta"
}