Data Mashups in R

Turning Up the Heat

PBSmapping allows us to see which in polygons/tracts our foreclosures were plotted. Using this data we can represent the intensity of foreclosure events as a heatmap.

  > addressPolys<-findPolys(addressEvents,myShapeFile)
  > addressPolys
       EID PID SID Bdry
   1   106   1   1    0
   2   218   3   1    0
   3   118   4   1    0
   4    40  13   1    0
   5   155  13   1    0
   6   294  14   1    0
   7   342  14   1    0
   8   361  14   1    0
   9   332  17   1    0
  10   343  17   1    0
  11   369  17   1    0
  (snip)

Factors When You Need Them

Each EID (event id aka foreclosure) is associated with a PID (polygon id aka tract). To plot our heatmap we need to count instances of PIDs in addressPolys for each tract on the map.

  > length(levels(as.factor(myShapeFile$PID))) 
  [1] 381

  > length(levels(as.factor(addressPolys$PID))) 
  [1] 196

We can see there are 381 census tracts in Philadelphia but only 196 have foreclosure events. For the purpose of coloring our polygons we need to insure remainder are explicitly set to 0 foreclosures.

The table function in R can be used to make this sort of contingency table. We need a variable, myTrtFC, to hold the number forclosures in each tract/PID:

  > myTrtFC<-
    table(factor(addressPolys$PID,levels=levels(as.factor(myShapeFile$PID))))
  > myTrtFC
  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20
  1  0  1  1  0  0  0  0  0  0  0  0  2  3  0  0  3  1  1  1
(snip)

(To enforce our new levels we must use a constructor (factor) instead of a variable conversion (as.factor))

Filling with Color Gradients

R has some decent built-in color gradient functions (?rainbow to learn more)—we ...