Databases
StackOverflow2013
Postsdbo
POIs there a good way to do this type of mining?
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POIs there a good way to do this type of mining?
    primarykey
    data
    text
    <p>I am trying to find points that are closest in space in X and Y directions (sample dataset given at the end) and am looking to see if there are smarter approaches to do this than my trivial (and untested) approach. The plot of these points in space looks something like the following and am trying to find sets of points marked inside the boxes i.e. the output I am looking for is a set of groups:</p> <pre><code>Group 1: (1,23), (2,23), (3,23)... Group 2: (68,200), (68,201), (68,203), (68,204), (68,100), (68,101), (68,101)... </code></pre> <p><img src="https://i.stack.imgur.com/YB5Eg.png" alt="enter image description here"></p> <p>For the horizontal bands, I am thinking I could just go ahead and use small sliding windows of size say, 5 or 10 (which should really be determined from the global information of which size will give the maximum grouped points but I am still exploring a good approach) and search for contiguous points because a break would not be considered a horizontal band anymore. </p> <p>I am guessing the same approach works for the vertical bands as well but not in all cases because there is a subtle difference in horizontal and vertical bands: points should appear close to be considered a group horizontally but they can appear anywhere to be considered part of a vertical band. Observe the large vertical band in the figure. So I am guessing I could just look for points that have the same x-coordinate (in this case, x=68) should give me a lot of points. </p> <p>Other than this trivial solution, I can't think of anything smart that can be done here as this problem appears deceptively simple to me. Am I missing something here? Does this fall into some known class of problems and if so, is there a good and scalable approach to achieve this?</p> <p><strong>Sample Dataset:</strong></p> <pre><code>1,23 1,23 2,23 3,23 4,23 5,23 6,23 7,23 8,23 9,23 10,23 11,23 12,23 13,23 14,23 15,23 16,23 10,33 11,33 12,33 13,33 14,33 15,33 16,33 17,33 18,33 19,33 2,28 2,28 3,28 34,75 34,76 34,76 34,77 34,78 34,79 34,80 34,81 34,82 34,83 34,75 34,76 34,76 34,77 34,78 34,79 34,80 34,81 400,28 400,28 400,28 68,200 68,201 68,203 68,204 68,100 68,101 68,103 68,104 </code></pre>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. This table or related slice is empty.
    1. USLegend
    plurals
    1. PL
      singulars
      1. LTLinked
    1. PL
      singulars
      1. LTLinked
    2. PL
      singulars
      1. LTLinked
    3. PL
      singulars
      1. LTLinked
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POIs there a good way to do this type of mining?
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POIs there a good way to do this type of mining?
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POIs there a good way to do this type of mining?
      1. USunutbu
      1. VTFavorite
    1. COClustering algorithm could be interesting ?
      singulars
      1. POIs there a good way to do this type of mining?
      1. USRicky Bobby
    2. COYou could try looking at the scipy.cluster module, though it looks like your system is a bit non-standard. Must the clusters be exclusively 1 row, 1 column, or are regions, e.g. (1,1), (1,2), (2,1), (2,2) allowed?
      singulars
      1. POIs there a good way to do this type of mining?
      1. USFredL
    3. COLook at R-tree algorithm
      singulars
      1. POIs there a good way to do this type of mining?
      1. USpawelzieba
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload