Merge pull request #48 from r-spatialecology/CRAN

Cran to master

R/data.R

@@ -11,12 +11,40 @@
 #'
 #' A species with negative associations to habitat 4 of `landscape`.
 #'
-#' @format A spatstat ppp object
+#' @format A spatstat ppp object.
 "species_a"
 
 #' Species b
 #'
 #' A species with positive associations to habitat 5 of `landscape`.
 #'
-#' @format A spatstat ppp object
+#' @format A spatstat ppp object.
 "species_b"
+
+#' Gamma test
+#'
+#' Randomized data for species b using the gamma test.
+#'
+#' @format rd_pat object.
+"gamma_test"
+
+#' Reconstruction
+#'
+#' Randomized data for species b using pattern reconstruction.
+#'
+#' @format rd_pat object.
+"reconstruction"
+
+#' Torus trans
+#'
+#' Torus translation of the classified landscape data.
+#'
+#' @format rd_ras object.
+"torus_trans"
+
+#' Random walk
+#'
+#' Randomization of the landscape data using the habitat randomization algorithm.
+#'
+#' @format rd_ras object.
+"random_walk"

R/plot_randomized_pattern.R

@@ -41,8 +41,6 @@ plot_randomized_pattern <- function(pattern,
                                     ask = TRUE,
                                     verbose = TRUE){
 
-
-
   # check if class is correct
   if (!class(pattern) %in% c("rd_pat", "rd_mar")) {
     stop("Class of 'pattern' must be 'rd_pat' or 'rd_mar'.",
@@ -232,6 +230,8 @@ plot_randomized_pattern <- function(pattern,
                        col = c("black", "#1f78b4"), lty = c(1,2), inset = 0.025)
 
       graphics::par(ask = FALSE)
+
+      invisible()
     }
 
     else if (class(pattern) == "rd_mar") {
@@ -306,11 +306,13 @@ plot_randomized_pattern <- function(pattern,
       graphics::legend(x = "topright",
                        legend = c("observed", "randomized"),
                        col = c("black", "#1f78b4"), lty = c(1, 2), inset = 0.025)
+
+      invisible()
     }
 
     else{
       stop("'method' must be either 'method = 'spatial' or 'method = 'marks'.",
-           call. =FALSE)
+           call. = FALSE)
     }
   }
 
@@ -368,6 +370,8 @@ plot_randomized_pattern <- function(pattern,
 
     # reset plotting window settings
     graphics::par(mfrow = c(1, 1))
+
+    invisible()
   }
 
   else{

R/plot_randomized_raster.R

@@ -141,4 +141,6 @@ plot_randomized_raster <- function(raster,
   # plot result
   raster::plot(raster_stack, col = col, nc = ncol, nr = nrow,
                colNA = "grey")
+
+  invisible()
 }

README.Rmd

@@ -6,6 +6,7 @@ output: github_document
 
 ```{r setup, include = FALSE}
 knitr::opts_chunk$set(
+  collapse = TRUE,
   fig.path = "man/figures/README-"
 )
 ```
@@ -52,10 +53,10 @@ landscape_classified <- classify_habitats(raster = landscape, classes = 5)
 
 There are two possibilities to randomize the environmental data, both described in Harms et al. (2001). The first shifts the habitat map in all 4 cardinal directions around a torus. The second one assigns the habitat values to an empty map using a random walk algorithm. Both functions return a list with randomized rasters and the observed one. For more information on the methods, please click [here](https://r-spatialecology.github.io/shar/articles/background.html).
 
-```{r habitat_random}
+```{r habitat_random, eval = FALSE}
 torus_trans <- translate_raster(raster = landscape_classified, verbose = FALSE)
 
-random_walk <- randomize_raster(raster = landscape_classified, n_random = 19, verbose = FALSE)
+random_walk <- randomize_raster(raster = landscape_classified, n_random = 39, verbose = FALSE)
 ```
 
 ```{r plot_habitat-random, fig.width = 5.5, fig.height = 4, fig.align = "center"}
@@ -66,28 +67,28 @@ plot_randomized_raster(torus_trans, n = 3, col = col)
 
 To randomize the point pattern, either use the Gamma test described by Plotkin et al. (2000) or pattern reconstruction (Tscheschel & Stoyan 2006). 
 
-```{r pattern-random}
-gamma_test <- fit_point_process(pattern = species_a, process = "cluster", n_random = 19, verbose = FALSE)
+```{r pattern-random, eval = FALSE}
+gamma_test <- fit_point_process(pattern = species_a, process = "cluster", n_random = 39, verbose = FALSE)
 
-reconstruct <- reconstruct_pattern(pattern = species_b, max_runs = 500, n_random = 19, verbose = FALSE) # takes some time
+reconstruction <- reconstruct_pattern(pattern = species_b, n_random = 39, verbose = FALSE) # takes some time
 ```
 
 Of coures, there are several utility functions. For example, you can plot a randomized pattern or calculate the differences between the observed pattern and the randomized patterns (using summary functions). 
 
 ```{r plot-random_pattern, fig.width = 5.5, fig.height = 4, fig.align = "center"}
-plot_randomized_pattern(reconstruct, verbose = FALSE, ask = FALSE)
+plot_randomized_pattern(reconstruction, verbose = FALSE, ask = FALSE)
 ```
 
 ```{r calculate-energy}
-calculate_energy(reconstruct, verbose = FALSE)
+calculate_energy(reconstruction, verbose = FALSE)
 ```
 
 The data was created that `species_a` has a negative association to habitat 4 and `species_b` has a positive association to habitat 5. At one point a posititive association to one habitat leads consequently to a negative association to another habitat (and vice versa). All this can be seen in the results.
 
 ```{r results}
 results_habitat_association(pattern = species_a, raster = torus_trans)
 
-results_habitat_association(pattern = reconstruct, raster = landscape_classified)
+results_habitat_association(pattern = reconstruction, raster = landscape_classified)
 ```
 
 ## References 

README.md

@@ -7,7 +7,7 @@ status](https://travis-ci.org/r-spatialecology/shar.svg?branch=master)](https://
 status](https://ci.appveyor.com/api/projects/status/08hgwkr82pqb6ykq/branch/master?svg=true)](https://ci.appveyor.com/project/mhesselbarth/shar/branch/master)
 [![Coverage
 status](https://codecov.io/gh/r-spatialecology/shar/branch/master/graph/badge.svg)](https://codecov.io/gh/r-spatialecology/shar?branch=master)
-[![Project Status: Active – The project has reached a stable, usable
+[![Project Status: Active – The project has reached a stable, usable
 state and is being actively
 developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
 [![lifecycle](https://img.shields.io/badge/lifecycle-maturing-blue.svg)](https://www.tidyverse.org/lifecycle/#maturing)
@@ -78,7 +78,7 @@ For more information on the methods, please click
 ``` r
 torus_trans <- translate_raster(raster = landscape_classified, verbose = FALSE)
 
-random_walk <- randomize_raster(raster = landscape_classified, n_random = 19, verbose = FALSE)
+random_walk <- randomize_raster(raster = landscape_classified, n_random = 39, verbose = FALSE)
 ```
 
 ``` r
@@ -94,30 +94,32 @@ Plotkin et al. (2000) or pattern reconstruction (Tscheschel & Stoyan
 2006).
 
 ``` r
-gamma_test <- fit_point_process(pattern = species_a, process = "cluster", n_random = 19, verbose = FALSE)
+gamma_test <- fit_point_process(pattern = species_a, process = "cluster", n_random = 39, verbose = FALSE)
 
-reconstruct <- reconstruct_pattern(pattern = species_b, max_runs = 500, n_random = 19, verbose = FALSE) # takes some time
+reconstruction <- reconstruct_pattern(pattern = species_b, n_random = 39, verbose = FALSE) # takes some time
 ```
 
 Of coures, there are several utility functions. For example, you can
 plot a randomized pattern or calculate the differences between the
 observed pattern and the randomized patterns (using summary functions).
 
 ``` r
-plot_randomized_pattern(reconstruct, verbose = FALSE, ask = FALSE)
+plot_randomized_pattern(reconstruction, verbose = FALSE, ask = FALSE)
 ```
 
 <img src="man/figures/README-plot-random_pattern-1.png" style="display: block; margin: auto;" /><img src="man/figures/README-plot-random_pattern-2.png" style="display: block; margin: auto;" />
 
 ``` r
-calculate_energy(reconstruct, verbose = FALSE)
+calculate_energy(reconstruction, verbose = FALSE)
+#>  [1] 0.01356462 0.01867389 0.01455831 0.01754614 0.01858267 0.02008615
+#>  [7] 0.01544740 0.01470591 0.01255234 0.01837727 0.01622154 0.01765028
+#> [13] 0.01517141 0.02075634 0.01836698 0.01523319 0.01832501 0.02072956
+#> [19] 0.01845929 0.01733616 0.01639753 0.01887276 0.01738685 0.01361427
+#> [25] 0.01619153 0.01695853 0.01856459 0.01692591 0.01986748 0.01604911
+#> [31] 0.01972613 0.01518272 0.01681647 0.02010567 0.01472619 0.01794340
+#> [37] 0.01740046 0.01531350 0.01869501
 ```
 
-    ##  [1] 0.06296249 0.04206978 0.05765356 0.04559935 0.07282586 0.05059367
-    ##  [7] 0.04891124 0.06394066 0.04618339 0.05240480 0.05701905 0.07678296
-    ## [13] 0.05384696 0.05210462 0.06433170 0.04546687 0.03842786 0.04996899
-    ## [19] 0.05361466
-
 The data was created that `species_a` has a negative association to
 habitat 4 and `species_b` has a positive association to habitat 5. At
 one point a posititive association to one habitat leads consequently to
@@ -126,30 +128,24 @@ be seen in the results.
 
 ``` r
 results_habitat_association(pattern = species_a, raster = torus_trans)
+#> > Input: randomized raster | Quantile thresholds: negative < 0.025 - positive > 0.975
+#>   habitat count lo hi significance
+#> 1       1     9  2 14         n.s.
+#> 2       2    25  9 24     positive
+#> 3       3    27 11 26     positive
+#> 4       4     0 11 29     negative
+#> 5       5    12  5 18         n.s.
+
+results_habitat_association(pattern = reconstruction, raster = landscape_classified)
+#> > Input: randomized point pattern | Quantile thresholds: negative < 0.025 - positive > 0.975
+#>   habitat count    lo    hi significance
+#> 1       1     8  4.85 43.15         n.s.
+#> 2       2    22 26.80 68.10     negative
+#> 3       3    33 46.80 74.10     negative
+#> 4       4    19 29.80 69.10     negative
+#> 5       5   118  9.95 37.05     positive
 ```
 
-    ## > Input: randomized raster | Quantile thresholds: negative < 0.025 - positive > 0.975
-
-    ##   habitat count lo hi significance
-    ## 1       1    10  0  8     positive
-    ## 2       2    14  8 24         n.s.
-    ## 3       3    30 14 29     positive
-    ## 4       4     0 10 26     negative
-    ## 5       5    14  4 17         n.s.
-
-``` r
-results_habitat_association(pattern = reconstruct, raster = landscape_classified)
-```
-
-    ## > Input: randomized point pattern | Quantile thresholds: negative < 0.025 - positive > 0.975
-
-    ##   habitat count    lo    hi significance
-    ## 1       1     7 10.45 26.75     negative
-    ## 2       2    20 31.35 49.10     negative
-    ## 3       3    31 50.45 73.55     negative
-    ## 4       4    33 36.45 59.65     negative
-    ## 5       5   109 23.45 43.65     positive
-
 ## References
 
 Baddeley, A., Rubak, E., Turner, R. (2015). Spatial Point Patterns:

data-raw/example_data.R

@@ -1,6 +1,7 @@
 library(dplyr)
 library(NLMR)
 library(maptools)
+library(usethis)
 library(spatstat)
 
 #### Create example data ####
@@ -19,32 +20,69 @@ landscape_class <- shar::classify_habitats(landscape, classes = 5)
 # Create species with negative
 pattern_a <- spatstat::runifpoint(n = 100, win = spatstat::owin(c(0, 1000), c(0, 1000)))
 
-owin_pattern <- raster::rasterToPolygons(landscape_class, fun = function(x){x == 4}, dissolve = TRUE) %>%
+owin_pattern <- raster::rasterToPolygons(landscape_class,
+                                         fun = function(x){x == 4},
+                                         dissolve = TRUE) %>%
   maptools::as.owin.SpatialPolygons()
 
 pattern_b <- pattern_a[!spatstat::inside.owin(x = pattern_a, w = owin_pattern)]
-pattern_c <- spatstat::rthin(pattern_a[spatstat::inside.owin(x = pattern_a, w = owin_pattern)], 0)
+pattern_c <- spatstat::rthin(pattern_a[spatstat::inside.owin(x = pattern_a,
+                                                             w = owin_pattern)], 0)
 
 species_a <- spatstat::superimpose(pattern_b, pattern_c,
                                    W = spatstat::owin(c(0, 1000), c(0, 1000)))
 
-marks_df_a <- data.frame(status = factor(sample(c("dead", "alive"), size = species_a$n, replace = TRUE)),
+marks_df_a <- data.frame(status = factor(sample(c("dead", "alive"),
+                                                size = species_a$n, replace = TRUE)),
                         dbh = runif(n = species_a$n, min = 5, max = 65))
 
 spatstat::marks(species_a) <- marks_df_a
 
 # Create species with positive associations
 pattern <- spatstat::runifpoint(n = 100, win = spatstat::owin(c(0, 1000), c(0, 1000)))
 
-species_b <- raster::rasterToPolygons(landscape_class, fun = function(x){x == 5}, dissolve = TRUE) %>%
+species_b <- raster::rasterToPolygons(landscape_class,
+                                      fun = function(x){x == 5}, dissolve = TRUE) %>%
   maptools::as.owin.SpatialPolygons() %>%
-  spatstat::runifpoint(n = floor(pattern$n * 1), win=.) %>%
+  spatstat::runifpoint(n = floor(pattern$n * 1), win = .) %>%
   spatstat::superimpose.ppp(pattern, W = spatstat::owin(c(0, 1000), c(0, 1000)))
 
-marks_df_b <- factor(sample(c("dominant", "understorey"), size = species_b$n, replace = TRUE))
+marks_df_b <- factor(sample(c("dominant", "understorey"),
+                            size = species_b$n, replace = TRUE))
 
 spatstat::marks(species_b) <- marks_df_b
 
-devtools::use_data(landscape, overwrite = TRUE)
-devtools::use_data(species_a, overwrite = TRUE)
-devtools::use_data(species_b, overwrite = TRUE)
+# translate raster
+torus_trans <- translate_raster(raster = landscape_class, verbose = FALSE)
+
+# use randomization algorithm
+random_walk <- randomize_raster(raster = landscape_class,
+                                n_random = 39, verbose = FALSE)
+
+# use gamma test
+gamma_test <- fit_point_process(pattern = species_b, process = "cluster",
+                                n_random = 39, verbose = FALSE)
+
+# use pattern reconstruction
+reconstruction <- reconstruct_pattern(pattern = species_b,
+                                      n_random = 39, verbose = FALSE)
+
+#### Save data ####
+
+# save landscape
+usethis::use_data(landscape, overwrite = TRUE)
+
+# save species data
+usethis::use_data(species_a, overwrite = TRUE)
+
+usethis::use_data(species_b, overwrite = TRUE)
+
+# save random landscape data
+usethis::use_data(torus_trans, overwrite = TRUE)
+
+usethis::use_data(random_walk, overwrite = TRUE)
+
+# save random point data
+usethis::use_data(gamma_test, overwrite = TRUE)
+
+usethis::use_data(reconstruction, overwrite = TRUE)