gdifazio / google-refine

Automatically exported from code.google.com/p/google-refine
0 stars 0 forks source link

Add support for a map facet #94

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago
Add a map facet to display geolocations.

The scatter facet already works similarly and so does most of the hard work 
already, but produces the output as an image.  An option next to the scatter 
facet 'show in map' might work, or we'd have to have a way to flag that a 
column is a coordinate.  (maybe could be could be inferred from the protograph?)

Original issue reported on code.google.com by iainsproat on 23 Jun 2010 at 4:48

GoogleCodeExporter commented 9 years ago
Some thoughts on how this might work (suggestions welcome):

A: the server getting a map graphic and adding it to the image produced by the 
scatter facet.  Returned and displayed in the UI like any other scatter facet.  
Some scatter facet features, such as 'Log', wouldn't be appropriate.

B: the server returning json output which is drawn in the client (like google 
maps or Open Street Map).  There may be problems with too much data here which 
needs some thought (clustering of data points by proximity etc..)

Original comment by iainsproat on 23 Jun 2010 at 4:51

GoogleCodeExporter commented 9 years ago
Issue 208 has been merged into this issue.

Original comment by iainsproat on 15 Nov 2010 at 10:08

GoogleCodeExporter commented 9 years ago

Original comment by tfmorris on 12 Dec 2010 at 7:49

GoogleCodeExporter commented 9 years ago
Since this ticket has been accepted (yay!!), some thoughts:

I would def. pursue option A.  The features that a complete client-side mapping 
implementation would offer (pan, alternate layers, etc.) aren't as pertinent in 
this case and it would likely make the ability to select a section of them 
significantly more complicated.  It should be feasible to fetch the Google 
tiles themselves (OpenLayers does this) and then crop and join the relevant 
sections to produce a map segment that will contain all the points.

From that point I would like to see it operate essentially like the scatterplot 
facet.  What I think is the most powerful possibility from my perspective is 
watching the map change as other facets are altered--being able to 
experimentally identify geographic patterns that correlate with features of the 
data.

Original comment by staringmonkey on 12 Dec 2010 at 11:29