boot_trees: Wrapper function for bootstrapped trees: reproducibly & in...
In mrajeev08/treerabid: Reconstruct Rabies Transmission Trees

boot_trees

R Documentation

Wrapper function for bootstrapped trees: reproducibly & in parallel

Description

Wrapper function for bootstrapped trees: reproducibly & in parallel

Usage

boot_trees(
  id_case,
  id_biter,
  y_coord,
  x_coord,
  owned,
  date_symptoms,
  days_uncertain,
  lineages = NULL,
  exclude_progen = FALSE,
  use_known_source = FALSE,
  prune = TRUE,
  si_fun,
  dist_fun,
  cutoff = 0.95,
  params,
  min_time = 1e-06,
  N = 1,
  seed = 1245,
  exp_funs = NULL,
  exp_pkgs = c("data.table", "treerabid", "igraph")
)

Arguments

`id_case`	id of case
`id_biter`	id of biter (i.e. known biter from contact tracing data, if unknown should be 0, if no contact tracing data, can pass NULL)
`y_coord`	y coordinate of case (should be in UTM: to do = use haversine distance for long/lat)
`x_coord`	x coordinate of case (should be in UTM: to do = use haversine distance for long/lat)
`owned`	whether animal is owned or not (per Katie, to account for uncertainty in case locations)
`date_symptoms`	case date (i.e. date symptoms started)
`days_uncertain`	uncertainty in days around case date
`lineages`	a data table with two columns, id_case and lineage, designating a lineage assignment for each case, defaults to NULL which means trees wont be resolved to a phylogeny
`exclude_progen`	boolean of length id_case or 1, if TRUE then case should be excluded as a potential progenitor (i.e. if including livestock cases or other species that are dead-end transmissions)
`use_known_source`	whether to assign known progenitors from contact tracing data
`prune`	whether to prune links (i.e. based on Cori et al./Mancy et al.) at a certain cutoff probability. This also results in assignment of incursions
`si_fun`	the function to get the probability of a given temporal difference between cases in days (i.e. the serial interval)
`dist_fun`	the function to get the probability of a given spatial difference between cases in meters (i.e. the dispersal kernel)
`cutoff`	the probability level at which to prune links (i.e. cases can not be linked if the distance or time difference is greater than the this %ile of the distribution); can either be a single number or a named vector with a cutoff called "dist" and "time"
`params`	list of parameters to pass to si_fun and dist_fun
`min_time`	if uncertainty results in negative time difference between known case pairs, set it to this value (better way to deal with propagating uncertainty given known case pairs?)
`N`	number of trees to build
`seed`	seed to pass to doRNG to make trees reproducible
`exp_funs`	functions to export to foreach loop (i.e. for customized si_fun/dist_fun which relies on an external scripts), may be needed for certain types of cluster configs
`exp_pkgs`	packages to export to foreach loop, defaults to data.table and treerabid, if other dependencies for si_fun or dist_fun, then pass here. May be needed for certain types of cluster configs