partition_data: Partition spatial data

In boyanangelov/sdmbench: Benchmark Species Distribution Models

Description

A function that partitions spatial data in order to avoid spatial autocorrelation.

Usage

partition_data(dataset_raster, dataset, env, method)

Arguments

dataset_raster	A raster dataset.
dataset	A dataframe containing species occurrences.
env	A raster dataset containing the environmental variables.
method	A character string representing the desired spatial partitioning method. Can be "default", "block", "checkerboard1", or "checkerboard2".

Value

A dataframe partitioned using the selected method.

Examples

## Not run: 
# download benchmarking data
benchmarking_data <- get_benchmarking_data("Lynx lynx",
                                           limit = 1500)

# apply data partitioning to benchmarking data
# note that this function overwrites the data
benchmarking_data$df_data <- partition_data(
     dataset_raster = benchmarking_data$raster_data,
     dataset = benchmarking_data$df_data,
     env = benchmarking_data$raster_data$climate_variables,
     method = "block")

# to inspect the partitioning results you can get a contingency table on the
# newly created grouping factor
# in this case you should see four different groups
table(benchmarking_data$df_data$label)

# use a different spatial partitioning method
benchmarking_data$df_data <- partition_data(
     dataset_raster = benchmarking_data$raster_data,
     dataset = benchmarking_data$df_data,
     env = benchmarking_data$raster_data$climate_variables,
     method = "checkerboard1")

# you can perform a sanity check here as well, you should see two groups
table(benchmarking_data$df_data$label)

## End(Not run)

boyanangelov/sdmbench documentation built on Dec. 14, 2020, 1:08 a.m.