Utilizing Rstudio as an alternative GIS - Revision history

Smitch: /* R and RStudio */

2023-10-30T17:54:46Z

R and RStudio

Smitch: /* QGIS */

2023-10-30T17:54:20Z

QGIS

Razzroutly: /* Install Packages */

2023-09-29T20:01:43Z

Install Packages

Razzroutly: /* Hillshade */

2023-09-29T19:58:34Z

Hillshade

Razzroutly: /* Hillshade */

2023-09-29T19:58:03Z

Hillshade

Razzroutly: /* Hillshade */

2023-09-29T19:57:48Z

Hillshade

Razzroutly: /* Laplacian Filter */

2023-09-29T19:50:11Z

Laplacian Filter

Razzroutly: /* Clipping */

2023-09-29T19:32:39Z

Clipping

Razzroutly: /* Raster to Vector */

2023-09-29T19:23:42Z

Raster to Vector

Razzroutly: /* Importing Data */

2023-09-29T19:22:34Z

Importing Data

@@ Line 32: / Line 32: @@
 === R and RStudio ===
-In order to utilize RStudio we first need to download '''R'''. This can be downloaded from https://cran.r-project.org/bin/windows/base/. For this installer we will be accepting all of the default paths and options.
+In order to utilize RStudio we first need to download '''R'''. This can be downloaded for Windows from https://cran.r-project.org/bin/windows/base/. For this installer we will be accepting all of the default paths and options.
 [[File:Rdownload 2023.png|600px|thumb|none| Figure 3: R 4.3.1 installer.]]

@@ Line 17: / Line 17: @@
 === QGIS ===
-We start by creating a fresh install of QGIS in order to download the functions that we will be exploring in this tutorial. QGIS can be installed from https://www.qgis.org/en/site/forusers/download.html, for the purposes of this tutorial we will be installing the 64-bit version.
+We start by creating a fresh install of QGIS in order to download the functions that we will be exploring in this tutorial. QGIS can be installed from https://www.qgis.org/en/site/forusers/download.html, for the purposes of this tutorial we will be installing the 64-bit version for Windows.  For other operating systems, follow directions at the QGIS website.
 [[File:QGISinstall.PNG|600px|thumb|none| Figure 1: QGIS 64-bit installer.]]

@@ Line 41: / Line 41: @@
 === Install Packages ===
 Start by opening up RStudio and creating a new ‘’’R script’’’ file (file->new file->R script). Within this new script we will be downloading and installing some packages that allow us to manipulate raster and spatial data within RStudio. Start by adding this code to your script and running it.
- '''#Install packages'''
+ #Install packages
  install.packages("sp")
- install.packages("RQGIS")
+ install.packages("sf")
- install.packages("rgeos")
+ install.packages("geodata")
+ install.packages("qgisprocess")
  install.packages("raster")
  install.packages("rgdal")
- '''#Attach packages to workspace'''
+ #Attach packages to workspace
  library("sp")
- library("RQGIS")
+ library("sf")
- library("rgeos")
+ library("geodata")
+ library("qgisprocess")
  library("raster")
  library("rgdal")
 *'''sp''': provides functions for plotting spatial data
+*'''sf''': provides functions for using spatial features
-*'''RQGIS''': Creates a link between the IDE (RStudio) and functions found within QGIS
+*'''geodata''': provides open source spatial data used in the tutorial
-*'''rgeos''': allows for the manipulation of vector data
+*'''qgisprocess''': Creates a link between the IDE (RStudio) and functions found within QGIS
 *'''raster''': Provides the user to read, write and manipulate raster images within RStudio
 *'''rgdal''': Provides the user to read, write and manipulate shapefiles within RStudio

@@ Line 155: / Line 155: @@
 == Hillshade ==
-The hillshade function creates a 3D representation of the original DEM. This function is available within RStudio and is called '''gdalogr:hillshade'''. In order to use this function we will need to set the azimuth angle and slope, we will use 270 and 40 degrees respectively. Hillshade analysis is useful for generating 3D images representative of the DEM plot. First we will create a hillshade raster with the '''gdalogr:hillshade''' algorithm, then we will overlay the original DEM to provide elevation as additional context to the image.
+The hillshade function creates a 3D representation of the original DEM. This function is available within RStudio and is called '''gdal:hillshade'''. In order to use this function we will need to set the azimuth angle and slope, we will use 270 and 40 degrees respectively. Hillshade analysis is useful for generating 3D images representative of the DEM plot. First we will create a hillshade raster with the '''gdal:hillshade''' algorithm, then we will overlay the original DEM to provide elevation as additional context to the image.
  #Hillshade function applied to the Canadian DEM, overlain by the original DEM
  qgis_run_algorithm("gdal:hillshade", INPUT=canada_dem_clip, AZIMUTH = 270, ALTITUDE=40, OUTPUT=paste(path,"/hillshade.tif", sep=""))

@@ Line 164: / Line 164: @@
 Below you can see the output generated from the hillshade as well as the output generated by overlaying the DEM.
-[[File:Hillshade BC.png|500px|left|thumb|none| Figure 12: Hillshade raster of the original Canadian DEM.]]
+[[File:Hillshade BC.png|500px|left|thumb|none| Figure 13: Hillshade raster of the original Canadian DEM.]]
-[[File:HillElev BC.png|500px|center|thumb|none| Figure 13: Canadian DEM overlaying the hillshade raster.]]
+[[File:HillElev BC.png|500px|center|thumb|none| Figure 14: Canadian DEM overlaying the hillshade raster.]]
 .

@@ Line 157: / Line 157: @@
 The hillshade function creates a 3D representation of the original DEM. This function is available within RStudio and is called '''gdalogr:hillshade'''. In order to use this function we will need to set the azimuth angle and slope, we will use 270 and 40 degrees respectively. Hillshade analysis is useful for generating 3D images representative of the DEM plot. First we will create a hillshade raster with the '''gdalogr:hillshade''' algorithm, then we will overlay the original DEM to provide elevation as additional context to the image.
  #Hillshade function applied to the Canadian DEM, overlain by the original DEM
- run_qgis(alg = "gdalogr:hillshade", INPUT = canada_dem_clip, AZIMUTH = 270,ALTITUDE=40,
+ qgis_run_algorithm("gdal:hillshade", INPUT=canada_dem_clip, AZIMUTH = 270, ALTITUDE=40, OUTPUT=paste(path,"/hillshade.tif", sep=""))
+ hillshade<-raster(paste(path,"/hillshade.tif", sep=""))
-          OUTPUT ='C://tutorial/canada_hillshadecan.tif')
+ plot(hillshade, col=grey(0:100/100), legend=FALSE, main='Canadian Hillshade and Elevation')
- canada_hillshade<-raster('C://tutorial/canada_hillshadecan.tif')
- plot(canada_hillshade, col=grey(0:100/100), legend=FALSE, main='Canadian Hillshade and Elevation')
  plot(canada_dem_clip, col=rainbow(25, alpha=0.35), add=TRUE)
 Below you can see the output generated from the hillshade as well as the output generated by overlaying the DEM.
-[[File:Hillshade.png|500px|left|thumb|none| Figure 15: Hillshade raster of the original Canadian DEM.]]
+[[File:Hillshade BC.png|500px|left|thumb|none| Figure 12: Hillshade raster of the original Canadian DEM.]]
-[[File:Hill elev.png|500px|center|thumb|none| Figure 16: Canadian DEM overlaying the hillshade raster.]]
+[[File:HillElev BC.png|500px|center|thumb|none| Figure 13: Canadian DEM overlaying the hillshade raster.]]
 .

@@ Line 145: / Line 145: @@
 == Laplacian Filter ==
-In this example we will utilize a Laplacian filter on the clipped version of the Canadian DEM. The Laplacian filter will be used to highlight the edges of the mountain range and gullies. To utilize a Laplacian filter within RStudio we will use a SAGA algorithm found within QGIS, '''saga:laplacianfilter'''. In order to use the Laplacian filter we need an input raster, kernel method, Radius, and sampling method. The code to run this filter on the clipped version of the Canadian DEM is provided below.
+In this example we will utilize a Laplacian filter on the clipped raster version of the Canadian DEM. The Laplacian filter will be used to highlight the edges of the mountain range and gullies. To utilize a Laplacian filter within RStudio we will use a SAGA algorithm found within QGIS, '''saga:laplacianfilter'''. In order to use the Laplacian filter we need an input raster, kernel method and type, radius, and sampling method. The code to run this filter on the clipped version of the Canadian DEM is provided below.
  #Creates a new raster image depicting the filtered Canadian DEM
- run_qgis(alg = "saga:laplacianfilter", INPUT = canada_dem_clip, METHOD = 2, MODE = 0, RADIUS = 1,
+ qgis_run_algorithm("saga:laplacianfilter", INPUT=canada_dem_clip, METHOD = 2, KERNEL_RADIUS = 1, KERNEL_TYPE = 0, RESULT=paste(path,"/laplacian.tif", sep=""))
+ lap_output<-raster(paste(path,"/laplacian.tif", sep=""))
-          SIGMA = 0,RESULT = 'C://tutorial/filter_dem.tif')
+ plot(lap_output,col=grey(0:100/100), main='Clipped Canada DEM')
- laplace<-raster('C://tutorial/filter_dem.tif')
- plot(laplace,col=grey(0:100/100),main = "Laplacian Filter")
 The resulting output is a filtered version of the original DEM. The Laplacian filter is often used for edge detection and is useful for analyzing mountain ranges as shown below.
-[[File:Laplacianfilter.png|500px|thumb|none| Figure 14: Canadian DEM filtered by the Laplacian algorithm.]]
+[[File:LaplacianFilter update.png|500px|thumb|none| Figure 12: Canadian DEM filtered by the Laplacian algorithm.]]
 == Hillshade ==

@@ Line 132: / Line 132: @@
 After running this code you will see a vector interpretation of the clipped study area. Here is a side by side comparison between the DEM and its shapefile counterpart.
-[[File:Canada Vectorized.png|500px|center|thumb|none| Figure 10: DEM of Canada study area displayed within the plot window.]]
+[[File:BC Vectorized.png|500px|center|thumb|none| Figure 10: DEM of British Columbia study area displayed within the plot window.]]
 === Vector clipping ===
-To clip the vector shapefile of the smaller Canadian study area that we created we will use the '''gdal:clipvectorbypolyon''' algorithm. This algorithm takes the shapefile as an input, a specified clipping extent, and a specified output for the result. Try running the code provided below.
+To demonstrate clipping vector data using qgisprocess we will now further clip down the British Columbia study area using the '''gdal:clipvectorbyextent''' algorithm. This algorithm takes the vector shapefile as an input, a specified clipping extent, and a specified output for the result. Try running the code provided below.
  '''#Take an input shapefile and clip it to the specified extent'''
+ poly_output <- qgis_run_algorithm("gdal:clipvectorbyextent", INPUT=canada_poly, EXTENT = "-121,-116,50.2,53.8")
- run_qgis(alg = "gdalogr:clipvectorsbyextent", INPUT_LAYER = shape, CLIP_EXTENT = '795592,797004,8933001,8934710', OUTPUT_LAYER = 'C://tutorial/vectclip')
+ vector_clip <- st_as_sf(poly_output)
- clipped_vect <- readOGR(dsn = "C://tutorial", layer = "vectclip")
- plot(clipped_vect,main='Shapefile of the clipped Mongón study area')
+ plot(vector_clip, main='Vectorized and Clipped DEM of study area')
-The result of the shapefile clip should display the shapefile that has been clipped to the extent specified within the '''run_qgis''' function.
+The result of the shapefile clip should display the shapefile that has been clipped to the extent specified within the '''qgis_run_algorithm''' function.
-[[File:Clipshape.png|500px|thumb|none| Figure 13: Clipped shapefile of the Mongón study area displayed within the plot window.]]
+[[File:SmallStudyArea.png|500px|thumb|none| Figure 11: Clipped shapefile of the study area displayed within the plot window.]]
 == Laplacian Filter ==

@@ Line 122: / Line 122: @@
 [[File:ClippedCanadaDEM.png|500px|thumb|none| Figure 9: Clipped DEM displayed within the plot window.]]
-== Raster to Vector ==
+=== Raster to Vector ===
 To demonstrate using clipping algorithms on vector files we will first need to change the raster of the Canada DEM to a vector using the GDAL '''polygonize''' algorithm which takes a raster image and converts its features into polygons. In order to use this function we will need to determine the parameters for: input, field, and output using the '''qgis_show_help()''' function. Since the output of this function is very large we will be storing the output vector in the temporary files of qgisprocess. This is done by leaving out the OUTPUT parameter causing it to default to qgis_temp_vector() and then calling the '''st_as_sf()''' function on the result of the algorithm.
  '''#Conversion of DEM to shapefile format'''

@@ Line 72: / Line 72: @@
 One of the advantages of using RStudio as an alternative GIS is its ability to load open source geospatial data with just one line of code. For this tutorial we will be using the '''geodata''' package to access digital elevation models using the elevation_30s() command which will provide a raster with 30 arc second (1km) from the Shuttle Radar Topography Mission, supplemented by GTOP30 data for high latitudes [https://rdrr.io/cran/geodata/man/elevation.html [1<nowiki>]</nowiki>].
-We will import the dataset that used in this tutorial into our workspace and display it within the RStudio '''plot''' frame. We will be downloading a DEM of Canada from the '''geodata''' package. The '''getData()''' function downloads data from an online database.
+We will import the dataset that used in this tutorial into our workspace and display it within the RStudio '''plot''' frame. We will be downloading a DEM of Canada from the '''geodata''' package. The '''getData()''' function downloads data from an online database. Due to the high resolution of the raster we will also use a function from the '''raster''' package that will aggregate the raster cells by a factor of 10 to make processing easier.
  '''#Digital elevation model (DEM) of Canada'''
  canada_dem<-getData('alt', country='CAN', mask=TRUE)
  plot(canada_dem, col=grey(0:100/100), main='DEM of Canada')
+ canada_dem_agg <- aggregate(canada_dem, fact=10)
 After running this code you will notice that the DEM's are displayed in the lower right '''plot''' window. The outputs should look like this:
 [[File:DEMCANADA1.png|500px|thumb|none| Figure 6: DEM of Canada displayed within the plot window.]]