Skip to contents

Sample points from a 2D environmental space potentially disjoint in geography

Source: R/selectionEG_helpers.R
point_sample_cluster.Rd

Sample one or more points from a two-dimensional environmental space according to a selection rule and with the possibility of having distinct sets of points to be sampled independently. Points to be sampled can be disjoint in geographic space and when that happens two points are selected considering the most numerous clusters.

Usage

point_sample_cluster(data, variable_1, variable_2, distance_list,
                     n = 1, cluster_method = "hierarchical",
                     select_point = "E_centroid", id_column = NULL)

Arguments

data

matrix or data.frame that contains at least four columns: "Longitude" and "Latitude" to represent geographic position, and two other columns to represent the variables of the 2D environmental space.

variable_1

(character or numeric) name or position of the first variable (x-axis).

variable_2

(character or numeric) name or position of the second variable (y-axis). Must be different from the first one.

distance_list

list of vectors of geographic distances among all points. If id_column is not defined, only one element in the list is needed, otherwise, distance_list must contain as many elements as unique IDs in id_column. In the latter case, the names in distance_list must match the IDs in id_column.

n

(numeric) number of points that are close to the centroid to be detected. Default = 1.

cluster_method

(character) there are two options available: "hierarchical" and "k-means". Default = "hierarchical".

select_point

(character) how or which point will be selected. Three options are available: "random", "E_centroid", and "G_centroid". E_ or G_ centroid indicate that the point(s) closest to the respective centroid will be selected. Default = "E_centroid".

id_column

(character or numeric) name or numeric index of the column in data containing identifiers of one or distinct sets of points. If, NULL, the default, only one set is assumed.

Value

A data.frame containing n rows corresponding to the point or points that were sampled.

Examples

# Data
m_matrix <- read_master(system.file("extdata/m_matrix.rds",
                                    package = "biosurvey"))
data("dist_list", package = "biosurvey")

# Making blocks for analysis
m_blocks <- make_blocks(m_matrix, variable_1 = "PC1", variable_2 = "PC2",
                        n_cols = 10, n_rows = 10, block_type = "equal_area")

datam <- m_blocks$data_matrix
datam <- datam[datam$Block %in% names(dist_list), ]

# Sampling points
point_clus <- point_sample_cluster(datam, variable_1 = "PC1",
                                   variable_2 = "PC2",
                                   distance_list = dist_list, n = 1,
                                   cluster_method = "hierarchical",
                                   select_point = "E_centroid",
                                   id_column = "Block")