Hi,
loessmodel <- loess(price ~ lat*lon, options) # fiddle with the options
locations <- expand.grid() # to make a prediction grid
averageprice <- predict(loessmodel, newdata = locations) # not a grid
locations$averageprice <- averageprice # to maka it a grid again
Or something along those lines. Of course there are also "better" ways to
interpolate.
Hope this helps,
Arien
Aleksandr Andreev wrote:
Hello folks!
I'm trying to create some nice graphs for summary statistics in a
paper. Here is the issue.
I have shapefiles of St Petersburg, Russia, which I can import via
spb<- readShapePoly("/home/sasha/Documents/maps/spb.shp")
I also have data on apartments in the city. For each apartment, I have
the latitude, longitude, and the price.
I would like to plot the price on the map in a "heat" style plot. The
natural way to do this by computing a local average price for a
specified neighborhood around (Lat, Lon) (say, of 0.01 of a degree
across -- I have enough data to make this as fine as I need).
Does anyone have suggestions on the best way to do this?
Thanks!
Aleks
---------------
Aleksandr Andreev
Graduate Student -- Department of Economics
University of North Carolina at Chapel Hill
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
--
Arien Lam
Dept. of Physical Geography
Utrecht University, NL
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
