lefko3 (version 6.2.0)

vrm_import: Import Vital Rate Model Factor Values for Function-based MPM Development

Description

Function vrm_import() builds a skeleton list holding data frames and vectors that can be used to import coefficient values for the factors of the vital rate models used to build function-based MPMs or run function-based projections.

Usage

vrm_import(
  years = NULL,
  patches = c(1),
  groups = c(0),
  interactions = FALSE,
  zi = FALSE,
  cat.indcova = NULL,
  cat.indcovb = NULL,
  cat.indcovc = NULL,
  dist.sizea = "gaussian",
  dist.sizeb = "constant",
  dist.sizec = "constant",
  dist.fec = "gaussian",
  trunc.sizea = FALSE,
  trunc.sizeb = FALSE,
  trunc.sizec = FALSE,
  trunc.fec = FALSE,
  use.juv = FALSE
)

Value

A list of class vrm_input, with up to 13 elements including:

vrm_frame

A data frame holding the main slope coefficients for the linear vital rate models.

year_frame

A data frame holding the main slope coefficients for the year at time t terms in the linear vital rate models.

patch_frame

A data frame holding the main slope coefficients for the patch terms in the linear vital rate models.

group2_frame

A data frame holding the main slope coefficients for the stage group terms in time t in the linear vital rate models.

group1_frame

A data frame holding the main slope coefficients for the stage group terms in time t-1 in the linear vital rate models.

dist_frame

A data frame giving the distributions of all variables, including primary, secondary, and tertiary size, and fecundity. Some variables begin as constant.

indcova2_frame

A data frame holding the main slope coefficients for the categorical individual covariate a terms in time t in the linear vital rate models.

indcova1_frame

A data frame holding the main slope coefficients for the categorical individual covariate a terms in time t-1 in the linear vital rate models.

indcovb2_frame

A data frame holding the main slope coefficients for the categorical individual covariate b terms in time t in the linear vital rate models.

indcovb1_frame

A data frame holding the main slope coefficients for the categorical individual covariate b terms in time t-1 in the linear vital rate models.

indcovc2_frame

A data frame holding the main slope coefficients for the categorical individual covariate c terms in time t in the linear vital rate models.

indcovc1_frame

A data frame holding the main slope coefficients for the categorical individual covariate c terms in time t-1 in the linear vital rate models.

st_frame

A data frame holding values of sigma or theta for use in Gaussian or negative binomial response terms, respectively.

The first element, called vrm_frame, is a data frame with the following 18 variables:

main_effect_1

The main effect for which coefficients are to be entered.

main_1_defined

A more natural explanation of main_effect_1.

main_effect_2

If given, then indicates another effect in a two-way interaction with main_effect_1.

main_2_defined

A more natural explanation of main_effect_2.

surv

A vector of coefficients for the factors in the model of adult survival.

obs

A vector of coefficients for the factors in the model of adult observation status.

sizea

A vector of coefficients for the factors in the model of adult primary size.

sizeb

A vector of coefficients for the factors in the model of adult secondary size.

sizec

A vector of coefficients for the factors in the model of adult tertiary size.

repst

A vector of coefficients for the factors in the model of adult reproductive status.

fec

A vector of coefficients for the factors in the model of adult fecundity.

jsurv

A vector of coefficients for the factors in the model of juvenile survival.

jobs

A vector of coefficients for the factors in the model of juvenile observation status.

jsize

A vector of coefficients for the factors in the model of juvenile primary size.

jsizeb

A vector of coefficients for the factors in the model of juvenile secondary size.

jsizec

A vector of coefficients for the factors in the model of juvenile tertiary size.

jrepst

A vector of coefficients for the factors in the model of juvenile reproductive status, for individuals maturing in the current time step.

jmat

A vector of coefficients for the factors in the model of maturity status, for individuals capable of maturing at the current time step.

sizea_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of adult primary size, if zero-inflated models are being used.

sizeb_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of adult secondary size, if zero-inflated models are being used.

sizec_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of adult tertiary size, if zero-inflated models are being used.

fec_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of fecundity, if zero-inflated models are being used.

jsizea_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of juvenile primary size, if zero-inflated models are being used.

jsizeb_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of juvenile secondary size, if zero-inflated models are being used.

jsizec_zi

A vector of coefficients for the factors in the binomial component of the zero-inflated model of juvenile tertiary size, if zero-inflated models are being used.

Arguments

years

A numeric vector of the years or times at time t to be modeled.

patches

A string or numeric vector of the patch names to be modeled.

groups

An integer vector of stage groups to be modeled. Defaults to a vector with a single element with value 0.

interactions

A logical value indicating whether to include two-way interactions between main effects (TRUE), or only main effects (FALSE). Defaults to FALSE.

zi

A logical value indicating whether to include coefficients for the binomial components of zero-inflation models. Defaults to FALSE.

cat.indcova

If individual covariate a is categorical, then this term should equal a string vector of the names of the categories. Defaults to NULL, in which case individual covariate a is either not used or is numeric.

cat.indcovb

If individual covariate b is categorical, then this term should equal a string vector of the names of the categories. Defaults to NULL, in which case individual covariate b is either not used or is numeric.

cat.indcovc

If individual covariate c is categorical, then this term should equal a string vector of the names of the categories. Defaults to NULL, in which case individual covariate c is either not used or is numeric.

dist.sizea

A string value giving the distribution of the variable coding primary size. Can equal "none", "gamma", "gaussian", "poisson", "negbin", or "constant". Defaults to "gaussian".

dist.sizeb

A string value giving the distribution of the variable coding secondary size. Can equal "none", "gamma", "gaussian", "poisson", "negbin", or "constant". Defaults to "constant".

dist.sizec

A string value giving the distribution of the variable coding tertiary size. Can equal "none", "gamma", "gaussian", "poisson", "negbin", or "constant". Defaults to "constant".

dist.fec

A string value giving the distribution of the variable coding fecundity. Can equal "none", "gamma", "gaussian", "poisson", or "negbin". Defaults to "gaussian".

trunc.sizea

A logical value indicating whether the distribution of the primary size variable should be zero-truncated. Defaults to FALSE. Currently only works with the Poisson and negative binomial distributions.

trunc.sizeb

A logical value indicating whether the distribution of the secondary size variable should be zero-truncated. Defaults to FALSE. Currently only works with the Poisson and negative binomial distributions.

trunc.sizec

A logical value indicating whether the distribution of the tertiary size variable should be zero-truncated. Defaults to FALSE. Currently only works with the Poisson and negative binomial distributions.

trunc.fec

A logical value indicating whether the distribution of the fecundity variable should be zero-truncated. Defaults to FALSE. Currently only works with the Poisson and negative binomial distributions.

use.juv

A logical value indicating whether to utilize juvenile vital rates. If FALSE, then all juvenile vital rates will be set to constant distributions. Defaults to FALSE.

Notes

All coefficients across all data frames are initially set to 0. After using this function to create the skeleton list, all relevant coefficient values should be set to non-zero values equal to the respective slope from the appropriate linear model, and any vital rate model to be used should have its distribution set to "binom", "gaussian", "gamma", "poisson", or "negbin". Unused vital rates should be set to "constant", and the first element of the correspoding column in vrm_frame (corresponding to the y-intercept) should be set to the constant value to utilize (generally 1). If no values are manually edited, then function-based MPM generator functions will not be able to generate valid MPMs.

Users should never change the labels or the order of terms in the data frames and vectors produced via this function, nor should they ever changes the names of core list elements in the vrm_input object. Doing so will result either in fatal errors or erroneous matrix calculations.

Using the vrm_input approach to building function-based MPMs requires careful attention to the stageframe. Although no hfv data frame needs to be entered, stages for which vital rates are to be estimated via linear models parameterized with coefficients provided via function vrm_import() should be marked as occurring within the dataset, while stages for which the provided coefficients should not be used should be marked as not occurring within the dataset.

Coefficients added to zero-inflation models can only be added to primary size, secondary size, tertiary size, fecundity, and the juvenile versions of primary, secondary, and tertiary size. Care must be taken to include zero- inflated coefficients only for variables without size-truncated distributions. Adding such terms will result in fatal errors during matrix creation.

Examples

data(lathyrus)

sizevector <- c(0, 100, 0, 1, 7100)
stagevector <- c("Sd", "Sdl", "Dorm", "ipm", "ipm")
repvector <- c(0, 0, 0, 1, 1)
obsvector <- c(0, 1, 0, 1, 1)
matvector <- c(0, 0, 1, 1, 1)
immvector <- c(1, 1, 0, 0, 0)
propvector <- c(1, 0, 0, 0, 0)
indataset <- c(0, 1, 1, 1, 1)
binvec <- c(0, 100, 0.5, 1, 1)
comments <- c("Dormant seed", "Seedling", "Dormant", "ipm adult stage",
  "ipm adult stage")
lathframeipm <- sf_create(sizes = sizevector, stagenames = stagevector, 
  repstatus = repvector, obsstatus = obsvector, propstatus = propvector,
  immstatus = immvector, matstatus = matvector, comments = comments,
  indataset = indataset, binhalfwidth = binvec, ipmbins = 100, roundsize = 3)

lathsupp2 <- supplemental(stage3 = c("Sd", "Sdl", "Sd", "Sdl"),
  stage2 = c("Sd", "Sd", "rep", "rep"),
  givenrate = c(0.345, 0.054, NA, NA),
  multiplier = c(NA, NA, 0.345, 0.054),
  type = c(1, 1, 3, 3), stageframe = lathframeipm, historical = FALSE)

lath_vrm <- vrm_import(years = c(1988:1990), zi = TRUE, dist.fec = "negbin",
  use.juv = TRUE)

lath_vrm$vrm_frame$surv[1] <- 2.32571
lath_vrm$vrm_frame$surv[2] <- 0.00109
lath_vrm$vrm_frame$obs[1] <- 2.230
lath_vrm$vrm_frame$sizea[1] <- 164.0695
lath_vrm$vrm_frame$sizea[2] <- 0.6211
lath_vrm$vrm_frame$fec[1] <- 1.517
lath_vrm$vrm_frame$fec_zi[1] <- 6.252765
lath_vrm$vrm_frame$fec_zi[2] <- -0.007313
lath_vrm$vrm_frame$jsurv[1] <- 1.03
lath_vrm$vrm_frame$jobs[1] <- 10.390
lath_vrm$vrm_frame$jsizea[1] <- 3.0559
lath_vrm$vrm_frame$jsizea[2] <- 0.8482

lath_vrm$year_frame$fec[c(1:3)] <- c(-0.41749627, 0.51421684, -0.07964038)
lath_vrm$year_frame$fec_zi[c(1:3)] <- c(3.741475e-07, -7.804715e-08,
  -2.533755e-07)
lath_vrm$year_frame$sizea[c(1:3)] <- c(96.3244, -240.8036, 144.4792)
lath_vrm$year_frame$jobs[c(1:3)] <- c(-0.7459843, 0.6118826, -0.9468618)
lath_vrm$year_frame$jsizea[c(1:3)] <- c(0.5937962, 1.4551236, -2.0489198)

lath_vrm$dist_frame$dist[2] <- "binom"
lath_vrm$dist_frame$dist[9] <- "binom"

lath_vrm$st_frame[3] <- 503.6167
lath_vrm$st_frame[7] <- 0.2342114
lath_vrm$st_frame[10] <- 5.831

lath_vrm$vrm_frame$sizeb[1] <- 1
lath_vrm$vrm_frame$sizec[1] <- 1
lath_vrm$vrm_frame$repst[1] <- 1
lath_vrm$vrm_frame$jsizeb[1] <- 1
lath_vrm$vrm_frame$jsizec[1] <- 1
lath_vrm$vrm_frame$jrepst[1] <- 1
lath_vrm$vrm_frame$jmatst[1] <- 1

lathmat2_importipm <- flefko2(stageframe = lathframeipm,
  modelsuite = lath_vrm, supplement = lathsupp2, reduce = FALSE)

summary(lathmat2_importipm)