lsm2multifit: Convert landscapeMetrics to Multifit format

In wilsonfrantine/landscapeDecoupler: Landscape Decoupler: Tools for Multi-scale Landscape Analysis

Convert landscapeMetrics to Multifit format

Description

this function was designed to shape the data as expected by multifit(), merging the outputs from calc_lsm() and some biological data to be used as reponse variables.

Usage

lsm2multifit(
  lsm,
  biodata = NULL,
  level = NULL,
  metrics = NULL,
  class = NULL,
  id.col = "site"
)

Arguments

lsm	a data frame from calc_lsm()
biodata	a dataframe with response variables; see euglossini example data frame
level	The level to extract the metric; must be "landscape" or "class".
metrics	The metric to be extracted from calc_lsm() output table
class	The class to extract the value if level is "class"
id.col	the column of the biodata that has the site id as the calc_lsm() output; "site" is default value

Value

a data frame in wide format with a single value of a given metrics for each scale

Author(s)

Wilson Frantine-Silva

Examples

## Not run: 
#For "landscape" level metrics
lsmdata <- calc_lsm(decouple(r,p,c(1000,2000,3000)), metric = "shdi")
d2multi <- lsm2multifit(lsm = lsmdata, biodata=euglossini,
level="landscape", metrics ="shdi")
head(d2multi)
scales <- colnames(d2multi)[grepl("X", colnames(d2multi))]
multifit(mod = "lm", multief = scales,
       formula = Abundance ~ multief, data = d2multi,
       criterion = "R2", plot_est = FALSE)

#For "class" level metrics
lsmdata <- calc_lsm(decouple(r,p,c(1000,2000,3000)), level = "class", metric = "pland")
biodata <- euglossini
d2multi <- lsm2multifit(lsm = lsmdata, biodata=euglossini, level="class", class=3, metrics ="shdi")
head(d2multi)
#' scales <- colnames(d2multi)[grepl("X", colnames(d2multi))]
multifit(mod = "lm", multief = colnames(d2multi)[grepl("X", colnames(d2multi))],
         formula = Abundance ~ multief, data = d2multi,
         criterion = "R2", plot_est = FALSE)

## End(Not run)

wilsonfrantine/landscapeDecoupler documentation built on Oct. 31, 2024, 3:45 a.m.