{
  "id": 8312,
  "title": "Any luck loading the .mat files into MNE?",
  "url": "/competitions/decoding-the-human-brain/discussion/8312",
  "author_name": "",
  "post_date": "2014-05-29T05:03:38.047Z",
  "votes": null,
  "comment_count": 3,
  "views": 1445,
  "content": "<p>On the main competition page it was mentioned that the python mne module was used to preprocess the data. In looking into the mne module, it looks like there are several powerful data exploration and manipulation features I would like to try out. Unfortunately, it seems that the raw file format accepted by the module is .fif files. Anyone been able to load the data into mne? Or can the competition creators offer any insight on how to reload the .mat files into mne?</p>",
  "messages": [
    {
      "id": "47224",
      "postDate": "05/29/2014 05:03:38",
      "content": "<p>On the main competition page it was mentioned that the python mne module was used to preprocess the data. In looking into the mne module, it looks like there are several powerful data exploration and manipulation features I would like to try out. Unfortunately, it seems that the raw file format accepted by the module is .fif files. Anyone been able to load the data into mne? Or can the competition creators offer any insight on how to reload the .mat files into mne?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "47286",
      "postDate": "05/29/2014 21:06:51",
      "content": "<p>have a look at the new EpochsArray class:</p>\n<p>https://github.com/mne-tools/mne-python/blob/master/mne/epochs.py#L1552</p>\n<p>it allows you to create an Epochs instance from a plain numpy array.</p>\n<p>You can also copy the learning code in mne/decoding and adapt it to your needs if it's simpler.</p>\n<p>hope this helps</p>\n<p>Alex</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "47382",
      "postDate": "05/31/2014 04:13:27",
      "content": "<p>Alex,</p>\n<p>Thank you very much for pointing me in this direction. I also looked at the .io.RawArray method. With RawArray I can successfully load the data, but it does not seem to play nice with the other MNE features. Also, I couldn't figure out how to load the data using EpochsArray. It seems that all my troubles in both cases are related to the 'events' variable that is often loaded directly from the .fif files. I don't have enough domain knowledge to know what 'events' is referring to, and it seems that a lot of the other methods requires this for successful parsing. Would you mind shedding some light on this variable?</p>\n<p>I really appreciate your insight,</p>\n<p>Josh</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "47397",
      "postDate": "05/31/2014 16:46:53",
      "content": "<p>with the competition data you need to work with EpochsArray (not RawArray)</p>\n<p>the events typically contains the info about the target y in the learning task.</p>\n<p>My feeling is that you should not care too much about what MNE implements</p>\n<p>as you'll need to go beyond in the context of the competition.</p>\n\n<p>Alex</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 47286,
      "author_name": "agramfort",
      "author_url": "",
      "post_date": "05/29/2014 21:06:51",
      "content": "<p>have a look at the new EpochsArray class:</p>\n<p>https://github.com/mne-tools/mne-python/blob/master/mne/epochs.py#L1552</p>\n<p>it allows you to create an Epochs instance from a plain numpy array.</p>\n<p>You can also copy the learning code in mne/decoding and adapt it to your needs if it's simpler.</p>\n<p>hope this helps</p>\n<p>Alex</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 47382,
      "author_name": "jstaker7",
      "author_url": "",
      "post_date": "05/31/2014 04:13:27",
      "content": "<p>Alex,</p>\n<p>Thank you very much for pointing me in this direction. I also looked at the .io.RawArray method. With RawArray I can successfully load the data, but it does not seem to play nice with the other MNE features. Also, I couldn't figure out how to load the data using EpochsArray. It seems that all my troubles in both cases are related to the 'events' variable that is often loaded directly from the .fif files. I don't have enough domain knowledge to know what 'events' is referring to, and it seems that a lot of the other methods requires this for successful parsing. Would you mind shedding some light on this variable?</p>\n<p>I really appreciate your insight,</p>\n<p>Josh</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 47397,
      "author_name": "agramfort",
      "author_url": "",
      "post_date": "05/31/2014 16:46:53",
      "content": "<p>with the competition data you need to work with EpochsArray (not RawArray)</p>\n<p>the events typically contains the info about the target y in the learning task.</p>\n<p>My feeling is that you should not care too much about what MNE implements</p>\n<p>as you'll need to go beyond in the context of the competition.</p>\n\n<p>Alex</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "47224": "",
    "47286": "",
    "47382": "",
    "47397": ""
  },
  "source": "meta"
}