bootstrap_glm: Bootstrap estimation of Generalized Linear Models on a matrix...
In surh/AMOR: Abundance Matrix Operations in R

bootstrap_glm

R Documentation

Bootstrap estimation of Generalized Linear Models on a matrix of observations

Description

Uses bootstrap to estimate variance of matrix_glm estimates. Takes a matrix of observations, and fits the specified GLM on each species. It perfomrs N bootstrap pesudoreplicates and returns the matrix of coefficients, for the original dataset as well as bootstrap estimates. It uses glm_matrix() to fit each of the pseudoreplicates.

Usage

bootstrap_glm(x, ...)

## Default S3 method:
bootstrap_glm(
  x,
  Map,
  formula,
  family = poisson(link = "log"),
  response.name = "Count",
  N = 100,
  verbose = FALSE,
  ...
)

## S3 method for class 'Dataset'
bootstrap_glm(
  x,
  formula,
  family = poisson(link = "log"),
  response.name = "Count",
  N = 100,
  verbose = FALSE,
  ...
)

Arguments

`x`	Either an abundance matrix with samples as columns or a Dataset object
`...`	Extra parameters for matrix_glm, which will in turn pass them to glm. Typically it is seful to specify the 'control' option.
`Map`	A mapping file with sample metadata if x is a matrix. A data.frame containing the variables to be modelled as columns and samples as rows. The rows should be named with sample IDs and must correspond to the column names from x if an abundance matrix was passed.
`formula`	The formula for the GLM model.Only the right hand side of the equation must be passed, and anything on the left side wild be silently ignored.
`family`	The model family. See help of glm and family for details
`response.name`	The name to give to the response variable
`N`	The number of bootstrap pseudoreplicates
`verbose`	Logical. If true each call to matrix_glm will print progress information.

Details

This works by calling matrix_glm() on the original data, and each pf the N bootstrap pseudoreplicates generated.

Value

A bootglm object which is a list containing the following elements:

orig: A matrix.glm object containing the output of matrix.glm() on the original data.
boot: A list containing elements: 1) coefficients, a three-dimensioanl array containing the matrix of coefficients for each bootstrap pseudoreplicate; and 2) SE, a three-dimensional array containing the matrix of standard errors for each bootstrap pseudoreplicate.
coefficients: A matrix of coefficients calculated as the mean accross each bootstrap pseudoreplicates.
call: Function call via match.call().
family: GLM model family. An object of class family.
pseudoreplicates: Matrix of dimensions (number iof samples) x (number of pseudoreplicates), where the i-column gives the sample ID, of the samples used for the i-th pseudoreplicate.
N: Number of bootstrap pseudoreplicates performed.