Skip to contents

[Deprecated]

Usage

splnr_convert_regionalisation(dat, PUs, cat_name = NA, col_name = NA)

Arguments

dat

Dataset in raster or sf format.

PUs

sf object of Planning Units

cat_name

A character string of all categories in the regionalisation

col_name

The name of the layer

Value

sf object containing the Planning Units and the feature.

Details

This is a wrapper for splnr_Convert2PUs() but deals with need to processes each layer seperately

The dataset needs to be raster or vector format. If the input contains continuous data, the output is an area-averaged mean for each planning unit. If the input is binary, the output is the proportion of the planning unit covered.

Examples

if (FALSE) { # \dontrun{
df <- splnr_convert_regionalisation(dat, PUs)
} # }