Skip to content

How to draw partial grid in plot for spatial-binomial experiment?

3 messages · Ruben Roa, Marc Schwartz (via MN), Roger Bivand

Original
Ruben Roa
# 
DeaR comRades:

I have a 2D spatial binomial process as shown in the data and code below.
I am plotting the number of trials and the number of successes in the spatial
binomial experiments and would like to draw the spatial cells were the trials
and successes were counted, i.e. a partial grid in the plot only for those 
cells where there is a number. The cells are 2x2 km cells. The count of
Trials and Success should ideally appear in the middle of the square cell. 
I know there is the 'grid' package but it seems the plots made using 'graphics'
are not compatible with the plots made using 'grid' (as warned in the grid help
pages). Thanks in advance.
Ruben

"fri"<-structure(list(
coords=structure(c(606,606,608,608,608,608,608,610,610,610,610,610,610,610,612,612,612,612,612,612,614,614,
614,614,614,614,614,614,614,616,616,616,616,616,616,616,618,618,618,618,618,620,620,620,622,624,
4388,4390,4384,4386,4388,4390,4392,4380,4382,4384,4386,4388,4390,4392,4380,4382,4384,4386,4388,4390,4374,
4376,4378,4380,4382,4384,4386,4388,4390,4372,4374,4376,4378,4380,4382,4384,4364,4366,4374,4376,4378,4368,
4374,4376,4366,4366),.Dim=c(46,2)),
data=c(3,2,0,0,11,4,0,1,1,3,5,9,3,0,0,16,7,0,0,0,0,0,0,0,4,1,0,0,0,0,4,9,12,0,0,0,0,0,4,5,2,1,0,0,0,0),
units.m=c(4,6,1,1,12,7,1,2,3,4,5,11,5,2,2,17,8,1,1,1,1,1,1,3,6,4,2,2,1,2,8,11,15,1,1,1,2,1,8,6,5,1,2,2,1,1),),
class="geodata")
par(mfrow=c(1,2))
plot(fri$coords[,1],fri$coords[,2],type="n",xlab="Easting (km)",ylab="Northing (km)",main="Success")
text(fri$coords[,1],fri$coords[,2],format(fri$data),cex=.6)
plot(fri$coords[,1],fri$coords[,2],type="n",xlab="Easting (km)",ylab="Northing (km)",main="Trials")
text(fri$coords[,1],fri$coords[,2],format(fri$units.m),cex=.6)
Marc Schwartz (via MN)
#
On Mon, 2005-12-19 at 11:17 -0200, Ruben Roa wrote:
Is this what you want?:


par(mfrow=c(1,2))

plot(fri$coords[,1],fri$coords[,2],type="n",
     xlab="Easting (km)",ylab="Northing (km)",
     main="Success")

text(fri$coords[,1],fri$coords[,2],format(fri$data),cex=.6)

# Use rect() to draw the grids around the values
# The sides of each rectangle will be +/- 1 from the 
# center point
rect(fri$coords[,1] - 1, fri$coords[,2] - 1, 
     fri$coords[,1] + 1 , fri$coords[,2] + 1)


plot(fri$coords[,1],fri$coords[,2],type="n",
     xlab="Easting (km)",ylab="Northing (km)",
     main="Trials")

text(fri$coords[,1],fri$coords[,2],format(fri$units.m),cex=.6)

# Same here
rect(fri$coords[,1] - 1, fri$coords[,2] - 1, 
     fri$coords[,1] + 1 , fri$coords[,2] + 1)


If so, see ?rect.

The 'grid' packages is the basis of the lattice graphics functionality.
It has nothing do to (directly) with drawing grid patterns on plots.

HTH,

Marc Schwartz
Roger Bivand
#
On Mon, 19 Dec 2005, Marc Schwartz (via MN) wrote:

            

      
      
      
    

        
An alternative is to use functions in the sp package:

library(sp)
fri2 <- SpatialPoints(fri$coords)
fri2_SP <- SpatialPixels(fri2)
fri2_SPl <- as.SpatialPolygons.SpatialPixels(fri2_SP)
fri2_SPl_df <- SpatialPolygonsDataFrame(fri2_SPl, 
  data.frame(data=fri$data, units.m=fri$units.m, 
  row.names=IDvaluesSpatialPixels(fri2_SP)))
opar <- par(mfrow=c(1,2))
plot(fri2_SPl_df, axes=TRUE)
text(coordinates(fri2_SPl_df), label=format(fri2_SPl_df$data), cex=0.6)
title(xlab="Easting (km)", ylab="Northing (km)", main="Success")
plot(fri2_SPl_df, axes=TRUE)
text(coordinates(fri2_SPl_df), label=format(fri2_SPl_df$units.m), cex=0.6)
title(xlab="Easting (km)", ylab="Northing (km)", main="Trials")
par(opar)

which captures most of the polygon drawing for you. It has to go from 
points to pixels to polygons to induce the regular shapes.

Roger