Does anyone know the equivalent of the spatial join tool in ArcMap that can be used in Hadoop?
I am trying to join point and polygon data in Hadoop and create a new point files that adds the attributes of the polygon data.
-
Edit the title to "in Hadoop" rather than "in ArcMap"?– Randall WhitmanApr 26, 2016 at 15:37
Add a comment
|
2 Answers
I think this would be spatial aggregation based on this article.
answered Apr 26, 2016 at 15:26
-
That article on spatial binning, plus the GIS Tools for Hadoop samples of Point-in-Polygon join, should provide the info needed to develop a spatial join. [collaborator] Apr 26, 2016 at 18:28
-
Geoffrey this would give me the count of points in a polygon. What i was looking for was returning the polygon each point is located within.– mkellerApr 27, 2016 at 20:30
Adapt the GIS Tools for Hadoop samples of Point-in-Polygon join, to output the desired info, rather than output counts.
Something to the effect of the following untested query:
SELECT earthquakes.*, counties.* from earthquakes, counties
WHERE ST_Contains(counties.boundaryshape, ST_Point(earthquakes.longitude, earthquakes.latitude));
[disclosure: collaborator]
-
I was able to use: SELECT eventid FROM earthquakes JOIN counties WHERE counties.name = 'San Bernardino' AND ST_Contains(counties.boundaryshape, ST_Point(earthquakes.longitude, earthquakes.latitude)); This gives me the earthquakes located within an individual county but i wanted to get a list of earthquakes for each county.– mkellerApr 29, 2016 at 13:38
-
Might be possible with Hive ARRAY - you could try looking up samples or reference documentation, and/or ask a separate question, about using ARRAY in HQL/SQL. Apr 29, 2016 at 15:09
-
1SELECT counties.name, earthquakes.eventid FROM counties JOIN earthquakes WHERE ST_Contains(counties.boundaryshape, ST_Point(earthquakes.longitude, earthquakes.latitude)) GROUP BY name, eventid This code let me list the polygon each point was in.– mkellerApr 29, 2016 at 16:05