prl_prepare_data: Prepare data for analysing with phurl models
In rdinnager/phurl: Phylogenetic modelling Using Regularized Linear models

Description Usage Arguments Value

View source: R/prepare_data.R

Prepare data for analysing with phurl models

1
2
3

prl_prepare_data(tree, traits = NULL, species_column = NULL,
  scale_edges = TRUE, centre = TRUE, transform_traits = NULL,
  standardise_traits = TRUE)

`tree`	Phylogenetic tree of class `phylo` or `tibble` (e.g. using `as_tibble`)
`traits`	Trait data. Can be a named `vector`, a `matrix` or `data.frame` with named rows, where the rownames are the species names, or a `data.frame` or `tibble` with a column specifying the species name. If species names are not referenced in either the rownames or a column, data will be matched to the tree using order only (e.g. assuming data rows are in the same order as `tree$tip.label`). Can also be `NULL` in which case only the predictor variables derived from the phylogeny are generated. This can be useful to feed into the simulation functions, or to use in a custom model somewhere else.
`species_column`	Character value specifying which column of `traits` refers to the species names, if they are not specified in the rownames. If not `NULL`, this will override any rownames in `traits`
`scale_edges`	Should the edges be scaled to a maximum of 1? This is highly recommended for regularised models. Default TRUE.
`centre`	Should the resulting edge path predictor matrix be centred to so that each column has a mean of 0? This is recommended to improve convergence of the models. Default TRUE.
`transform_traits`	List of transformations to apply to trait values. If named, will be matched to column names, else they will be applied in order, and recycled if necessary.
`standardise_traits`	Should traits values be standardised by subtracting the mean and dividing by them standard deviation? Recommended. Default TRUE.