Exercise 1_4_1: Basic Raster Operations

1

Create a matrix with 8 rows and 8 columns filled with random numbers between 1 and 1000.

Clues

You can create the numbers using the sample() function.

solution

library(dplyr)

fancy_matrix <-
  sample(1:1000, 8*8, replace = TRUE) |> 
  matrix(nrow = 8, ncol = 8)

fancy_matrix

##      [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8]
## [1,]  752   27  933  578  174  948  974  604
## [2,]   89  462   49  837  574  788  108  829
## [3,]  502  500  863  862  531  875  358  890
## [4,]  960  365  793  841  408  974  539  697
## [5,]  130  541  808  961  981  484  986  143
## [6,]   38  576  110  839  532  294  764   35
## [7,]  784  424  768  718   85  727  838   27
## [8,]    3  216  183  858  849  805  192  929

2

Now, use this matrix to create a raster layer and plot it.

Clues

The terra::rast() function can be fed with matrices to create a raster layer.

solution

library(terra)

fancy_raster_layer <- terra::rast(fancy_matrix)

terra::plot(fancy_raster_layer)

The terra::rast() function can not only be used to create raster data on the fly, which is also quite boring. Instead, we can use it to import already prepared data.

3

Import one of the raster .tiff files in the ./data folder of the workshop directory.

Clues

Make sure your file paths are set correctly. You can check them with getwd(). Setting is done with setwd().

solution

immigrants_cologne <- terra::rast("./data/immigrants_cologne.tif")

4

Import the data on immigrants and inhabitants. Create a new layer showing the proportion of immigrants in each grid cell between 0 and 1. Next, apply z-standardization and dichotomize the data to get information about which grid cells deviate from the overall mean. Plot both “versions” of the data to see how they compare.

Clues

You can handle raster layers as any simple data table using +, -, or / operators. Z-standardization can be applied using the terra::scale() function.

solution

# load all layers
immigrants_cologne <-
  terra::rast("./data/immigrants_cologne.tif")

inhabitants_cologne <-
  terra::rast("./data/inhabitants_cologne.tif")

# create proportation layer
immigrants_proportion <- immigrants_cologne / inhabitants_cologne

# scale data
immigrants_proportion_scaled <- terra::scale(immigrants_proportation)

immigrants_proportion_scaled[immigrants_proportion_scaled < 0] <- 0
immigrants_proportion_scaled[immigrants_proportion_scaled > 0] <- 1

library(tmap)

tm_shape(immigrants_proportion) +
  tm_raster(palette = "-viridis")

tm_shape(immigrants_proportion_scaled) +
  tm_raster(palette = "-viridis")