I have household data, which includes GPS coordinates and distance to a point (let this var be distkm) for that household. Some of the households are missing this information. I thought about applying the same methodology that is used to find the epicenter of earthquakes. You take three points, form a circle of radius distkm, then when three radii intersect, you've located your point. This may go awry as I know that the point of interest varies, but I'm assuming households that are close refer to the same point (perhaps finding all possible intersections?). Additionally, the coordinates and distkm may be slightly off, so there would have to be some tolerance for error. Has anyone done anything like this in Stata or am I looking at GIS type software? Thanks!
Here's some randomly generated data, to at least act as an example:
Here's some randomly generated data, to at least act as an example:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input str15 lat str11 lat_1 str16 longi str11 longi_1 str8 distkm "42°27′42″N" "42.46162604" "115°55′04″E" "115.9176455" "957.0455" "38°48′38″N" "38.81061172" "116°26′16″E" "116.4376907" "553.3505" "33°26′13″N" "33.43706894" "126°41′51″E" "126.6974838" "784.6264" "25°20′33″N" "25.3423747" "117°09′42″E" "117.1616089" "" "39°01′41″N" "39.02797534" "120°44′40″E" "120.7444089" "596.7802" "32°48′35″N" "32.80966886" "109°01′37″E" "109.0268566" "" "32°25′01″N" "32.41696514" "127°56′10″E" "127.9360959" "919.4911" "35°58′04″N" "35.96777766" "118°48′46″E" "118.8127057" "" "41°12′02″N" "41.20064759" "113°27′40″E" "113.4610452" "899.4456" "34°48′31″N" "34.80852712" "124°29′32″E" "124.4921518" "579.2791" end
