A wrapper for all the crest functions.

Runs all the different steps of a CREST reconstruction in one function.

crest(
  df,
  climate,
  pse = NA,
  taxaType = 0,
  distributions = NA,
  site_info = rep(NA, length(climate)),
  site_name = NA,
  xmn = NA,
  xmx = NA,
  ymn = NA,
  ymx = NA,
  continents = NA,
  countries = NA,
  realms = NA,
  biomes = NA,
  ecoregions = NA,
  minGridCells = 20,
  selectedTaxa = NA,
  bin_width = rep(1, length(x$parameters$climate)),
  shape = rep("normal", length(x$parameters$climate)),
  npoints = 500,
  ai.sqrt = FALSE,
  geoWeighting = TRUE,
  climateSpaceWeighting = TRUE,
  climateSpaceWeighting.type = "linear",
  presenceThreshold = 0,
  taxWeight = "normalisation",
  uncertainties = c(0.5, 0.95),
  leave_one_out = FALSE,
  verbose = TRUE,
  dbname = "gbif4crest_02"
)

Arguments

df: A data frame containing the data to reconstruct (counts, percentages or presence/absence data).
climate: A vector of the climate variables to extract. See accClimateVariables for the list of accepted values.
pse: A pollen-Species equivalency table. See createPSE for details.
taxaType: A numerical index (between 1 and 6) to define the type of palaeoproxy used: 1 for plants, 2 for beetles, 3 for chironomids, 4 for foraminifers, 5 for diatoms and 6 for rodents. The example dataset uses taxaType=0 (pseudo-data). Default is 1.
distributions: A dataframe containing the presence records of the studied proxies and their associated climate values.
site_info: A vector containing the coordinates of the study site. Default c(NA, NA).
site_name: The name of the dataset (default NA).
xmn, xmx, ymn, ymx: The coordinates defining the study area.
continents: A vector of the continent names defining the study area.
countries: A vector of the country names defining the study area.
realms: A vector of the studied botanical realms defining the study area.
biomes: A vector of the studied botanical biomes defining the study area.
ecoregions: A vector of the studied botanical ecoregions defining the study area.
minGridCells: The minimum number of unique presence data necessary to estimate a species' climate response. Default is 20.
selectedTaxa: A data frame assigns which taxa should be used for each variable (1 if the taxon should be used, 0 otherwise). The colnames should be the climate variables' names and the rownames the taxa names. Default is 1 for all taxa and all variables.
bin_width: The width of the bins used to correct for unbalanced climate state. Use values that split the studied climate gradient in 15-25 classes (e.g. 2°C for temperature variables). Default is 1.
shape: The imposed shape of the species pdfs. We recommend using 'normal' for temperature variables and 'lognormal' for the variables that can only take positive values, such as precipitation or aridity. Default is 'normal' for all.
npoints: The number of points to be used to fit the pdfs. Default 200.
ai.sqrt: A boolean to indicate whether ai values should be square-root transformed (default FALSE).
geoWeighting: A boolean to indicate if the species should be weighting by the square root of their extension when estimating a genus/family level taxon-climate relationships.
climateSpaceWeighting: A boolean to indicate if the species pdfs should be corrected for the modern distribution of the climate space (default TRUE).
climateSpaceWeighting.type: A correction factor for the clame space weighting correction to limit the edge effects. Either 'linear' (default), 'sqrt' or 'log'.
presenceThreshold: All values above that threshold will be used in the reconstruction (e.g. if set at 1, all percentages below 1 will be set to 0 and the associated presences discarded). Default is 0.
taxWeight: One value among the following: 'originalData', 'presence/absence', 'percentages' or 'normalisation' (default).
uncertainties: A (vector of) threshold value(s) indicating the error bars that should be calculated (default both 50 and 95% ranges).
leave_one_out: A boolean to indicate whether the leave one out (loo) reconstructions should be computed (default FALSE).
verbose: A boolean to print non-essential comments on the terminal (default TRUE).
dbname: The name of the database. Default is 'gbif4crest_02'.

Value

A crestObj containing the reconstructions.

Examples

if (FALSE) {
  data(crest_ex)
  data(crest_ex_pse)
  data(crest_ex_selection)
  reconstr <- crest(
    df = crest_ex, pse = crest_ex_pse, taxaType = 0,
    site_info = c(7.5, 7.5), site_name = 'crest_example',
    climate = c("bio1", "bio12"), bin_width = c(2, 50),
    shape = c("normal", "lognormal"),
    selectedTaxa = crest_ex_selection, dbname = "crest_example",
    leave_one_out = TRUE,
    verbose = FALSE
  )
  plot(reconstr)
  plot_loo(reconstr)
}