address spell check using R CMD CHECK and regenerate pkgdown

docs/articles/process-landscape.html

@@ -235,7 +235,7 @@ <h2 id="remotely-sensed-land-cover">1. Remotely sensed land cover<a class="ancho
 NDVI is a measure of healthy green vegetation, based on the tendency of
 plants to reflect NIR &amp; absorb red light. It ranges from -1
 (non-vegetated) to 1 (densely vegetated). The figure below shows the
-histogram of NDVI values withiin Punggol, as well as the adaptively
+histogram of NDVI values within Punggol, as well as the adaptively
 derived threshold value (vertical line).</p>
 <div class="sourceCode" id="cb8"><pre class="downlit sourceCode r">
 <code class="sourceCode R"><span class="va">ndvi_mosaic</span> <span class="op">&lt;-</span> 

docs/pkgdown.yml

@@ -13,7 +13,7 @@ articles:
   process-animal: process-animal.html
   process-landscape: process-landscape.html
   use-cases: use-cases.html
-last_built: 2022-10-28T07:34Z
+last_built: 2022-10-28T07:57Z
 urls:
   reference: https://ecological-cities.github.io/biodivercity/reference
   article: https://ecological-cities.github.io/biodivercity/articles

docs/sitemap.xml

@@ -133,7 +133,7 @@
     <loc>https://ecological-cities.github.io/biodivercity/reference/sampling_points.html</loc>
   </url>
   <url>
-    <loc>https://ecological-cities.github.io/biodivercity/reference/standard_SR.html</loc>
+    <loc>https://ecological-cities.github.io/biodivercity/reference/standard_sr.html</loc>
   </url>
   <url>
     <loc>https://ecological-cities.github.io/biodivercity/reference/tally_observations.html</loc>

inst/WORDLIST

@@ -66,3 +66,38 @@ ebird
 Hirundinidae
 Hirundo
 tahitica
+etc
+Brassica
+GLMM
+GLMMs
+Giam
+HTW
+KY
+Kurukulasuriya
+PCNM
+NIR
+Neto
+Peres
+Przeslawski
+RDA
+Radke
+Rajathurai
+Siwabessy
+Teo
+Tran
+YF
+biodiversity'
+decompose'
+etc
+gamma'
+geostatistical
+gui
+knitr
+landscapemetrics
+performance'
+png
+rustica
+spp
+subzones
+ΔAIC
+β

vignettes/build-models.Rmd

@@ -206,7 +206,7 @@ rm(birds, birds_scaled, predictors, subset_exp,
 
 ## 2. Community (_Beta_) diversity
 
-Partial RDA models were used to model the relationship between animal communities and landscape predictors while considering also the inherent spatial structure of the animal communities. Similarly, variable selection by random forest was performed on the landscape variables prior to beta model building. In the following sections, the example variables were selected using the `randomForestSRC` and `MultivariateRandomForest` packages. After the first round of variable selection, the landscape predictors were summerised as Principal Component Analysis (PCA) axes while the spatial variation were quantified using Principle Coordinates of Neighbourhood Matrices (PCNM).
+Partial RDA models were used to model the relationship between animal communities and landscape predictors while considering also the inherent spatial structure of the animal communities. Similarly, variable selection by random forest was performed on the landscape variables prior to beta model building. In the following sections, the example variables were selected using the `randomForestSRC` and `MultivariateRandomForest` packages. After the first round of variable selection, the landscape predictors were summarised as Principal Component Analysis (PCA) axes while the spatial variation were quantified using Principle Coordinates of Neighbourhood Matrices (PCNM).
 
 First, load the additional package `vegan` for beta diversity model building:
 
@@ -330,15 +330,15 @@ birds.rda.fin <- rda(bird_com ~ PC1 + PC2 + PC4 + PC6 + PC7 + PC8 + Condition(PC
 
 ```
 
-The predictive β-diversity model also requires information on the average probability of each species occurring across all points.
+The predictive _Beta_-diversity model also requires information on the average probability of each species occurring across all points.
 
 ```{r p-hats}
 # fitted_126 <- fitted(birds1.rda.fin, model = "CCA")
 p_hats_birds <- apply(bird_com, 2, mean)
 p_hats_birds
 ```
 
-The PCA axes, PNCM vectors, final RDA model and probability of species occurence are saved out to be applied in the subsequent apply model section.
+The PCA axes, PCNM vectors, final RDA model and probability of species occurrence are saved out to be applied in the subsequent apply model section.
 
 ```{r save beta model objs, eval = FALSE}
 save(pca_birds,

vignettes/more/animals-simulate-missing-surveys.Rmd

@@ -64,7 +64,7 @@ knitr::kable(head(scenarios), caption = "**First few rows of the object `scenari
 
 <br>
 
-## Visualise reuslts
+## Visualise results
 
 These 100 bootstrapped scenarios may be used in further analyses, or visualised to observe the variation in species richness within the sampled area of interest (Punggol): 
 

vignettes/process-landscape.Rmd

@@ -127,7 +127,7 @@ mosaic_sen2r(parent_dir = paste0(output_dir, "/sen2r"))
 
 <br>
 
-At this point, we have a single image mosaic (continuous raster) for each spectral index, for a given area/period of interest. While the continuous values from these rasters may be used directly in analyses, there may be instances were we want to work with discrete classes of land cover. To do so, we can classify pixels into one of two classes (e.g. vegetated or non-vegetated; water or non-water), based on an adaptively derived threshold value. For example, we can use Otsu's thresholding ([Otsu, 1979](https://cw.fel.cvut.cz/wiki/_media/courses/a6m33bio/otsu.pdf)), which tends to outperform other techniques in terms of stability of results and processing speed, even with the presence of > 2 peaks in the histogram of pixel values ([Bouhennache et al., 2019](https://www.tandfonline.com/doi/abs/10.1080/10106049.2018.1497094)). This can be implemented using the function `threshold_otsu()`. As an example, let's load the NDVI raster for the Punggol area in Singapore for subsequent classification. The NDVI is a measure of healthy green vegetation, based on the tendency of plants to reflect NIR & absorb red light. It ranges from -1 (non-vegetated) to 1 (densely vegetated). The figure below shows the histogram of NDVI values withiin Punggol, as well as the adaptively derived threshold value (vertical line).
+At this point, we have a single image mosaic (continuous raster) for each spectral index, for a given area/period of interest. While the continuous values from these rasters may be used directly in analyses, there may be instances were we want to work with discrete classes of land cover. To do so, we can classify pixels into one of two classes (e.g. vegetated or non-vegetated; water or non-water), based on an adaptively derived threshold value. For example, we can use Otsu's thresholding ([Otsu, 1979](https://cw.fel.cvut.cz/wiki/_media/courses/a6m33bio/otsu.pdf)), which tends to outperform other techniques in terms of stability of results and processing speed, even with the presence of > 2 peaks in the histogram of pixel values ([Bouhennache et al., 2019](https://www.tandfonline.com/doi/abs/10.1080/10106049.2018.1497094)). This can be implemented using the function `threshold_otsu()`. As an example, let's load the NDVI raster for the Punggol area in Singapore for subsequent classification. The NDVI is a measure of healthy green vegetation, based on the tendency of plants to reflect NIR & absorb red light. It ranges from -1 (non-vegetated) to 1 (densely vegetated). The figure below shows the histogram of NDVI values within Punggol, as well as the adaptively derived threshold value (vertical line).
 
 ```{r}
 ndvi_mosaic <-