Raster Prediction with factors in model

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Raster Prediction with factors in model

Mark Payne-2
Hi,

I have a GAM model that contains factors in addition to smooth terms etc. I
would like to evaluate this at all points on a raster, but I can't seem to
get the predict.raster() function to pick up the values of my factors. I
have created a minimum (not) working example here, based on the example for
the predict() function.

Any suggestions in this direction would be appreciated.

Mark

# create a RasterStack or RasterBrick with with a set of predictor layers
logo <- brick(system.file("external/rlogo.grd", package="raster"))

# known presence and absence points
p <- matrix(c(48, 48, 48, 53, 50, 46, 54, 70, 84, 85, 74, 84, 95, 85,
              66, 42, 26, 4, 19, 17, 7, 14, 26, 29, 39, 45, 51, 56, 46, 38,
31,
              22, 34, 60, 70, 73, 63, 46, 43, 28), ncol=2)

a <- matrix(c(22, 33, 64, 85, 92, 94, 59, 27, 30, 64, 60, 33, 31, 9,
              99, 67, 15, 5, 4, 30, 8, 37, 42, 27, 19, 69, 60, 73, 3, 5, 21,
              37, 52, 70, 74, 9, 13, 4, 17, 47), ncol=2)

# extract values for points
xy <- rbind(cbind(1, p), cbind(0, a))
v <- data.frame(cbind(xy[,1], extract(logo, xy[,2:3])))
colnames(v)[1] <- 'pa'

#Convert blue layer to a factor
v$blue <- factor(v$blue < 200)

#build a model, here an example with glm
model <- glm(formula=pa~., data=v)

#Setup prediction objects
pred.b <- logo[[1:2]]
const.df <- data.frame(blue=TRUE)

#predict to a raster
r1 <- predict(pred.b, model, progress='text',
              const=const.df)


#Returns an error as follows:
# Error in `[.data.frame`(blockvals, , f[j]) : undefined columns selected

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Raster Prediction with factors in model

Robert Hijmans
Mark,
I had not considered the case where the constant is a factor. I need
to fix that. Here is a script that accomplishes what you want, I think

# known presence and absence points
p <- matrix(c(48, 48, 48, 53, 50, 46, 54, 70, 84, 85, 74, 84, 95, 85,
              66, 42, 26, 4, 19, 17, 7, 14, 26, 29, 39, 45, 51, 56, 46, 38,31,
              22, 34, 60, 70, 73, 63, 46, 43, 28), ncol=2)

a <- matrix(c(22, 33, 64, 85, 92, 94, 59, 27, 30, 64, 60, 33, 31, 9,
              99, 67, 15, 5, 4, 30, 8, 37, 42, 27, 19, 69, 60, 73, 3, 5, 21,
              37, 52, 70, 74, 9, 13, 4, 17, 47), ncol=2)

# extract values for points
xy <- rbind(cbind(1, p), cbind(0, a))
v <- data.frame(cbind(xy[,1], extract(logo, xy[,2:3])))
colnames(v)[1] <- 'pa'

#Convert blue layer to a factor
## RH I would not use TRUE/FALSE but an integer representation instead
v$blue <- factor((v$blue < 200) + 0)

#build a model, here an example with glm
model <- glm(formula=pa~., data=v)

#Setup prediction objects
pred.b <- logo
pred.b[[3]] <- (pred.b[[3]] < 200) + 0
# or all TRUE?  pred.b[[3]] <- 1
names(pred.b) <- names(logo)

#predict to a raster
r1 <- predict(pred.b, model, progress='text')


You say you want to use a gam, but the example is a glm. With gam
there might be additional issues. Please let me know if that is the
case.

Robert


On Mon, Apr 20, 2015 at 3:47 AM, Mark Payne <[hidden email]> wrote:

> Hi,
>
> I have a GAM model that contains factors in addition to smooth terms etc. I
> would like to evaluate this at all points on a raster, but I can't seem to
> get the predict.raster() function to pick up the values of my factors. I
> have created a minimum (not) working example here, based on the example for
> the predict() function.
>
> Any suggestions in this direction would be appreciated.
>
> Mark
>
> # create a RasterStack or RasterBrick with with a set of predictor layers
> logo <- brick(system.file("external/rlogo.grd", package="raster"))
>
> # known presence and absence points
> p <- matrix(c(48, 48, 48, 53, 50, 46, 54, 70, 84, 85, 74, 84, 95, 85,
>               66, 42, 26, 4, 19, 17, 7, 14, 26, 29, 39, 45, 51, 56, 46, 38,
> 31,
>               22, 34, 60, 70, 73, 63, 46, 43, 28), ncol=2)
>
> a <- matrix(c(22, 33, 64, 85, 92, 94, 59, 27, 30, 64, 60, 33, 31, 9,
>               99, 67, 15, 5, 4, 30, 8, 37, 42, 27, 19, 69, 60, 73, 3, 5, 21,
>               37, 52, 70, 74, 9, 13, 4, 17, 47), ncol=2)
>
> # extract values for points
> xy <- rbind(cbind(1, p), cbind(0, a))
> v <- data.frame(cbind(xy[,1], extract(logo, xy[,2:3])))
> colnames(v)[1] <- 'pa'
>
> #Convert blue layer to a factor
> v$blue <- factor(v$blue < 200)
>
> #build a model, here an example with glm
> model <- glm(formula=pa~., data=v)
>
> #Setup prediction objects
> pred.b <- logo[[1:2]]
> const.df <- data.frame(blue=TRUE)
>
> #predict to a raster
> r1 <- predict(pred.b, model, progress='text',
>               const=const.df)
>
>
> #Returns an error as follows:
> # Error in `[.data.frame`(blockvals, , f[j]) : undefined columns selected
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo