{
  "id": 21535,
  "title": "Spatial Features and Visualisation on a Map",
  "url": "/competitions/grupo-bimbo-inventory-demand/discussion/21535",
  "author_name": "NikolayBurlutskiy",
  "post_date": "2016-06-09T10:36:25.760000",
  "votes": 4,
  "comment_count": 0,
  "views": 1025,
  "content": "<p>I wrote <a href=\"https://www.kaggle.com/nikolayburlutskiy/grupo-bimbo-inventory-demand/spatial-features-and-visualisation\">a script</a> in Python for the following:</p>\n\n<p>1) Summarising the frequency of sales depot towns to appear in the train dataset. The more times products were sold from a town means that the more frequent that town is;</p>\n\n<p>2) Extracting spatial information on these towns such as their latitude, longitude, importance, type. I used GeoPy for geocoding;</p>\n\n<p>3) Visualising the towns as circles on a map where the size of the circles is proportional to the frequency of  sales depot towns to appear in the train dataset. </p>\n\n<p>Feel free to reuse and explore the features further. Geocoding doesn't work 100% correctly but I believe it can help to find some spatial patterns in product sales.</p>",
  "messages": [
    {
      "id": 123043,
      "postDate": "2016-06-09T10:36:25.760Z",
      "content": "<p>I wrote <a href=\"https://www.kaggle.com/nikolayburlutskiy/grupo-bimbo-inventory-demand/spatial-features-and-visualisation\">a script</a> in Python for the following:</p>\n\n<p>1) Summarising the frequency of sales depot towns to appear in the train dataset. The more times products were sold from a town means that the more frequent that town is;</p>\n\n<p>2) Extracting spatial information on these towns such as their latitude, longitude, importance, type. I used GeoPy for geocoding;</p>\n\n<p>3) Visualising the towns as circles on a map where the size of the circles is proportional to the frequency of  sales depot towns to appear in the train dataset. </p>\n\n<p>Feel free to reuse and explore the features further. Geocoding doesn't work 100% correctly but I believe it can help to find some spatial patterns in product sales.</p>",
      "rawMarkdown": "I wrote [a script][1] in Python for the following:\r\n\r\n1) Summarising the frequency of sales depot towns to appear in the train dataset. The more times products were sold from a town means that the more frequent that town is;\r\n\r\n2) Extracting spatial information on these towns such as their latitude, longitude, importance, type. I used GeoPy for geocoding;\r\n\r\n3) Visualising the towns as circles on a map where the size of the circles is proportional to the frequency of  sales depot towns to appear in the train dataset. \r\n\r\nFeel free to reuse and explore the features further. Geocoding doesn't work 100% correctly but I believe it can help to find some spatial patterns in product sales.\r\n\r\n  [1]: https://www.kaggle.com/nikolayburlutskiy/grupo-bimbo-inventory-demand/spatial-features-and-visualisation",
      "votes": 4
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "123043": "I wrote [a script][1] in Python for the following:\r\n\r\n1) Summarising the frequency of sales depot towns to appear in the train dataset. The more times products were sold from a town means that the more frequent that town is;\r\n\r\n2) Extracting spatial information on these towns such as their latitude, longitude, importance, type. I used GeoPy for geocoding;\r\n\r\n3) Visualising the towns as circles on a map where the size of the circles is proportional to the frequency of  sales depot towns to appear in the train dataset. \r\n\r\nFeel free to reuse and explore the features further. Geocoding doesn't work 100% correctly but I believe it can help to find some spatial patterns in product sales.\r\n\r\n  [1]: https://www.kaggle.com/nikolayburlutskiy/grupo-bimbo-inventory-demand/spatial-features-and-visualisation"
  }
}