updated raster template post

Author: wenzmo

_posts/rastertemplate/rastertemplate.Rmd

@@ -1,7 +1,8 @@
 ---
 title: "Raster Operations"
 description: |
-  Learn how to obtain a template raster from the `{d6geodata}` package, expanding the raster in different ways, rasterizing vector data, and extracting information from a raster using vector data. The methods described involve the use of packages like `{terra}` and `{sf}`, and include examples and plots to illustrate the processes. Additionally, we show the use of the {tidyterra} package for plotting raster data with ggplot.
+  Learn how to 1) obtain a template raster from the `{d6geodata}` package, 2) expand the raster in different ways, 3) rasterize vector data, and 4) extract information from a raster using vector data. 
+  The methods described involve the use of packages like `{terra}` and `{sf}`, and include examples and plots to illustrate the processes. Additionally, we show how to use the `{tidyterra}` package for plotting raster data with `{ggplot2}`.
   
 preview: ./img/wiki/preview-rastertemplate.png
 categories:
@@ -47,7 +48,7 @@ theme_set(
 
 ## Raster Template {#template}
 
-First get the template raster. The origin is the extent of Berlin with the projection laea (EPSG:3035).
+First get the template raster. The origin is the extent of Berlin with the projection LAEA (EPSG:3035).
 
 ```{r template}
 temp <- d6geodata::temp_ras(1000) # resolution is 1000m
@@ -62,37 +63,41 @@ We load Berlin districts data set from our cloud to show the borders of Berlin.
 berlin <- d6geodata::get_geodata(data_name = "districs_berlin_2022_poly_03035_gpkg",
                                  path_to_cloud = "E:/PopDynCloud",
                                  download_data = FALSE) %>% 
-  st_union() %>% # using `st_union()` to combine  
+  st_union() %>% # using `st_union()` to combine geometries
   st_cast("POLYGON")
 
 p_berlin <- geom_sf(data = berlin, alpha = 0, color = "black", lwd = 0.7)
 ```
 
+Now we plot the map.
 
 ```{r plot template}
 p1 <- d6geodata::plot_qualitative_map(temp) # you can use this build-in plot function from the `{d6geodata}` package 
  
 p1 + p_berlin
 ```
 
+
 ## Expand Raster {#expand}
 
 ### Expand On All Sides
 
-If you want to expand the raster on all size (similar to a buffer around a polygon) you can just add cells to all sides. You can do this by adding an integer number in the y argument in the function `extend` of the `{terra}` package.
+If you want to expand the raster (similar to a buffer around a polygon) you can just add cells to all sides. You can do this by adding an integer number in the y argument in the function `extend` of the `{terra}` package.
 
 ```{r}
+# expand the map
 temp_exp_all <- terra::extend(x = temp, # template raster as input
                               y = 10, # number of cells to add
                               fill = 2) # value for new cells
 
+# plot the map again
 p2 <- d6geodata::plot_qualitative_map(temp_exp_all)
 p2 + p_berlin
 ```
 
 ### Expand On Two Sides
 
-In this case you can add cells on two opposite sides.
+You can also expand the raster in one or two different directions only by adding cells on two opposite sides.
 
 ```{r}
 temp_exp_tb_sides <- terra::extend(x = temp,
@@ -102,9 +107,9 @@ p3 <- d6geodata::plot_qualitative_map(temp_exp_tb_sides)
 p3 + p_berlin
 ```
 
-### Expand On One Side With A Costum Extent
+### Expand On One Side With A Custom Extent
 
-In this case we use a workaround by using a costum extent. We are adding on the left (western) side cells by using the xmin of the larger extend. Xmax, ymin, ymax are from the given extent.
+In this case we use a workaround by using a custom extent. Here we are adding cells on the left (western) side by using the *xmin* of the larger extend. Xmax, ymin, ymax are from the given extent.
 
 ```{r}
 temp_exp_1_sides <- terra::extend(x = temp,
@@ -123,19 +128,22 @@ p4 + p_berlin
 Here you see a combination of all plots made to compare them directly.
 
 ```{r, preview = TRUE}
+# arrange plots in a 2x2 grid
 (p1 + p_berlin + theme(legend.position="none") + p2 + p_berlin) / (p3 + p_berlin + p4 + p_berlin) + 
   plot_layout(guides = "collect") +
-   plot_annotation(tag_levels = "1", tag_prefix = "(P", tag_suffix = ")")
+   plot_annotation(tag_levels = "1", tag_prefix = "(P", tag_suffix = ")") # add tag to every plot
 ```
 
 ## Rasterize Vector Data {#rasterize}
 
-With `vect()` you can make a spatVector (a kind of shapfile) out of an extent object.
+With `vect()` you can make a spatVector (similar to a shapefile) from an extent object.
 Afterwards, we can use this spatVector to rasterize the data. This works similar with an sf object.
 
 ```{r plot}
-temp_vec <- terra::vect(ext(temp), crs(temp)) # spatVector out of smalest raster extent
+# create spatVector
+temp_vec <- terra::vect(ext(temp), crs(temp)) # spatVector out of smallest raster extent
 
+# rasterize spatVector
 temp_rast <- rasterize(temp_vec, # vector data
                        temp_exp_all %>% disagg(10), 
                        # raster data with larger extent, disaggregated to 100m 
@@ -144,28 +152,34 @@ temp_rast <- rasterize(temp_vec, # vector data
 temp_rast
 ```
 
+
 ### Rasterized Plot
 
+View the results.
+
 ```{r}
 d6geodata::plot_qualitative_map(temp_rast)
 ```
 
 ## Extract With Vector Data {#extract}
 
-Imagine you have a track of your animal and you want to see, which type of information lays in this track. In this case, you can use your line and extract the values from a raster below. There are several ways to this but the easiest way is to use the `extract()` function form the `{terra}` package to get the values laying under the vector data.
+Imagine you have a track of an  animal and you want to know what information lays in this track. In this case, you can use your line and extract the values from a raster below. There are several ways to this but the easiest way is to use the `extract()` function form the `{terra}` package to get the values laying under the vector data.
 
 For this, you can use any vector data (points, lines, polygons) with the same projection and extent to extract values from a raster. The `extract()` function uses the raster as first input and the vector data as second. Depending on the type of data you are using you can define a function for the extraction like min, max or mean. 'Cells' gives you the call ID, 'xy' the coordinates and 'ID' the row number of your extraction.
+As mentioned, it is important that both the vector and the raster have the same projection.
 
 ### Line For Extraction
 
-For this we just creating a dummy line with coordinates laying within the raster
+For this we are just creating a dummy line with coordinates laying within the raster
 
 ```{r}
+# create points
 pts <- tibble(x = c(4528790, 4576890), # a point out of two coordinates from the raster
               y = c(3271740, 3272140))
 
+# create line from points
 dummy_line <- st_as_sf(pts,             # a line made out of the point data
-                       coords = c("x", "y"),
+                       coords = c("x", "y"), # specify columns of xy coordinates
                        crs = st_crs(temp)$wkt, # using the same projection as for the template raster
                        sf_column_name = "geometry",
                        remove = FALSE) %>% 
@@ -177,8 +191,8 @@ dummy_line <- st_as_sf(pts,             # a line made out of the point data
 ### Plot Of Raster And Line
 
 ```{r}
-p5 <- d6geodata::plot_qualitative_map(temp_rast) + 
-  geom_sf(data = dummy_line)
+p5 <- d6geodata::plot_qualitative_map(temp_rast) + # first plot raster
+  geom_sf(data = dummy_line)                       # plot line on top of the raster
 
 p5
 
@@ -193,31 +207,31 @@ ext_tab <- terra::extract(x = temp_rast, # template raster as first input
                y = dummy_line, # dummy line as second input
                fun=NULL, # here you can set a function to summarise the data directly
                method="simple", 
-               cells=FALSE, # here you can get the cell number as well
-               xy=FALSE, # here the xy coordinates 
+               cells=FALSE, # here you can get the cell number as well if cell = TRUE
+               xy=FALSE, # here the xy coordinates if xy = TRUE
                ID=TRUE # and the id in front of each row
                )
 ```
 
-As you already can see on the plot, the largest part of the line lays within the purple area (1). This function can be used in cases like you need to know in which landuse class your track lays.
+As you already can see on the plot, the largest part of the line lays within the purple area (value = 1). This function can be used to know on which land use class your track lays, for example.
 
 ```{r}
 table(ext_tab$layer)
 ```
 
-## Visualize data with tidyterra {#tidyterra}
+## Visualize Data With Tidyterra {#tidyterra}
 
-There are several ways to plot raster data with ggplot and one relatively new way is to do it with the package `{tidyterra}`. It has some well developed `{ggplot}` functions, but can sometimes not be combined with the base functions of `{ggplot}`. If you use this pacakge you may have to stay within the package functions.
+There are several ways to plot raster data with ggplot and one relatively new way is to do it with the package `{tidyterra}`. It has some well developed `{ggplot}` functions, but it cannot always be combined with the base functions of `{ggplot}`. If you use this package you may have to stay within the package functions.
 
 
 ```{r}
 library(tidyterra)
 
 ggplot() + 
-  geom_spatraster(data = temp_rast) +
-  p_berlin
-
+  geom_spatraster(data = temp_rast) + # use geom_spatraster() to plot the raster data with {tidyterra}
+  p_berlin # add vector data of Berlin
 
+# same a before, but add some color to the plot
 ggplot() + 
   geom_spatraster(data = temp_rast) +
   scale_fill_whitebox_c(palette = "purple") +