CODEWITHSUNDEEP
rspatialraster
Antti
Antti

Reputation: 1293

How to rasterize point data by weighted average in R?

I have a data.frame of spatial data with crop yields by field. I want to make a heat map by rasterizing this data. I have already been able to produce a map by using raster mean values. I could improve upon this by calculating weighted averages instead, the weight being the plot area. Thus far I haven't been able to make this work.

This is what I've done as a minimal example:

require(dplyr)
require(raster)
require(sf)

d <- data.frame(longitude = runif(100, 10, 30),
                latitude = runif(100, 40, 50),
                yield = runif(100, 2, 6),
                area = runif(100, .1, 10)) %>%
    st_as_sf(coords = c("longitude", "latitude"))

r <- raster(extent(c(min(d$longitude), max(d$longitude),
                     min(d$latitude), max(d$latitude))))

x <- rasterize(d, r, field = "yield", mean)

In order to achieve what I need I tried e.g. replacing the last line with:

x <- rasterize(d, r, field = "yield", function(y){weighted.mean(y, w = area)})

but that does not work as I thought. Any idea how to do this?

Upvotes: 1

Views: 478

Answers (1)

Majid
Majid

Reputation: 1854

I have modified what you did in creating a reproducible example to avoid getting error. Hope this helps:

#require(dplyr) #looks redundant
require(raster)
require(sf)

d <- data.frame(longitude = runif(100, 10, 30),
                latitude = runif(100, 40, 50),
                yield = runif(100, 2, 6),
                area = runif(100, .1, 10)) %>%
                st_as_sf(coords = c("longitude", "latitude"))

r <- raster(extent(c(st_bbox(d)))) #modified

x1 <- rasterize(d, r, field = "yield", mean)
x2 <- rasterize(d, r, field = "yield", fun=function(x,...) (sum(x[1]*x[2])/sum(x[2])))

Upvotes: 1

Related Questions