Skip to contents

Multiscale Heterogenity Map

Source: R/MHM.R
MHM.Rd

Description

Usage

MHM(
  x,
  window = window_quick(),
  kernel = kernel_circle,
  fun = mean,
  na.policy = "omit",
  fillvalue = NA,
  expand = FALSE,
  na.rm = TRUE,
  ...
)

Arguments

x

SpatRaster

window

numeric sequence of window sizes (default to c(3, 13, 23) via window_quick). Passed to kernel function

kernel

function among kernels (default to kernel_circle)

fun

function that takes multiple numbers, and returns a numeric vector (one or multiple numbers). For example mean, modal, min or max

na.policy

character. Can be used to determine the cells of x for which focal values should be computed. Must be one of "all" (compute for all cells), "only" (only for cells that are NA) or "omit" (skip cells that are NA). Note that the value of this argument does not affect which cells around each focal cell are included in the computations (use na.rm=TRUE to ignore cells that are NA for that)

fillvalue

numeric. The value of the cells in the virtual rows and columns outside of the raster

expand

logical. If TRUE The value of the cells in the virtual rows and columns outside of the raster are set to be the same as the value on the border. Only available for "build-in" funs such as mean, sum, min and max

na.rm

logical passed to fun. Whether to remove NA in the calculation for each focal cell. Not the NA in the global SpatRaster. See terra::focal

...

additional arguments passed to fun such as na.rm

Value

SpatRaster

References

Gaucherel, C. (2007) Multiscale heterogeneity map and associated scaling profile for landscape analysis, Landscape and Urban Planning 82(3) 95-102. doi: 10.1016/j.landurbplan.2007.01.022

Examples

# load terra
library(terra)

# import an example file
# and turn it into a SpatRaster
landscape <- import_example("l1.tif")

# plot it
plot(landscape)


# make it smaller for example purpose:
l0 <- landscape %>% raster_resample(0.1)

# plot the little landscape
p(l0)


# calculate MHM on it
l0_mhm <- MHM(l0,
              window=seq(2, 13, by=2)+1,
              fun=shannon_evenness)

# display all monoscale maps
p(l0_mhm)


# the profile plot
ms_profile(l0_mhm)

ms_profile(l0_mhm, summary_fun=median, error_fun=sd, ylab="median +/- sd")


# and a synthetic map
app(l0_mhm,  median) %>% p(title="multiscale map (median)")

app(l0_mhm,  sd) %>% p(title="sd", palette="Plasma")