Skip to contents

Generates a randomized version of a phylospatial object by extracting the tip community matrix, permuting it using nullcat::quantize(), and rebuilding the phylospatial object using the permuted tip matrix.

Usage

ps_quantize(ps, site_groups = NULL, wt_row = NULL, wt_col = NULL, ...)

Arguments

ps

Object of class phylospatial

site_groups

Optional vector giving each site's group, so that sites are randomized independently within groups. One element per site (as for site_weight in phylospatial()) or per occupied site. See ps_rand() for details.

wt_row, wt_col

No longer supported, and must be NULL (the default); supplying a weight matrix gives an error. Use site_groups instead. See ps_rand() for details.

...

Additional arguments passed to quantize, such as method, n_strata, transform, fixed, n_iter, etc.

Value

A randomized version of ps

Details

The nullcat quantize routine involves three steps: converting a quantitative matrix to categorical strata, permuting the resulting categorical matrix using one of several categorical null model algorithms, and mapping the randomized categories back to quantitative values. Supply arguments via ... to control options for each of these stages.

For repeated randomizations to generate a null distribution, it is more efficient to use ps_rand(fun = "quantize"), which is structured to avoid unnecessarily recomputing overhead that is shared across randomizations.

See also

Examples

# \donttest{
if (requireNamespace("nullcat", quietly = TRUE)) {
  ps <- ps_simulate(data_type = "prob")
  ps_rand <- ps_quantize(ps, n_strata = 4,
    n_iter = 1000,
    method = "curvecat", fixed = "cell")
}
#> Warning: 216 of 396 rows and 2 of 10 columns have all their values in a single stratum, so the randomization cannot change them under `fixed = "cell"`: their randomized values will equal the observed values. To randomize them, separate low values into their own strata (e.g. `zero_stratum = TRUE`, or a `transform`), or use a different `fixed` mode.
# }