DOC.boot: Bootstraps for loess, negative slope and fns

In Russel88/DOC: DOC analysis of microbial communities

Bootstraps for loess, negative slope and fns

Description

Bootstraps for loess, negative slope and fns

Usage

DOC.boot(do, R = 100, subr = NULL, pair = NULL, mov.avg = 5,
  span = 0.2, degree = 1, family = "symmetric", iterations = 4,
  surface = "interpolate", cores = 1)

Arguments

do	Output from DOC.do function
R	Number of bootstraps
subr	If NULL will do bootstrap, alternatively an integer denoting size of subsample
pair	A vector of length two with names of pairs
mov.avg	Moving average window to use for estimating where negative slope starts
span	Span of loess smoothing
degree	Degree of loess smoothing (If 1 linear, >1 polynomial)
family	"gaussian" is least-squares fitting, "symmetric" is robust fitting
iterations	Number of iterations for robust fitting
surface	"direct" estimation (slow exact) or "interpolate" estimation (fast approximate)
cores	Number of cores to use

Value

Output: A list with:

• A dataframe of loess for each bootstrap,

• a numeric with estimated slope from lme on data subsetted to Overlap values above median for each bootstrap,

• a numeric with estimated point of where negative slope starts (moving average of loess used) for each bootstrap,

• and a numeric with estimated Fns value for each bootstrap

Russel88/DOC documentation built on Nov. 14, 2022, 3:50 p.m.