do_sim: Run a simulation of a given scenario, fit models, report...

In dill/ltdesigntester: Test Performance of Line Transect Designs

Description

Given a simulation scenario, this function will generate from that specification, fit models to the data and record the results.

Usage

do_sim(nsim, scenario, pred_dat, stratification = c(), logit_opts = NULL,
  transect_id = NULL, cov_values = NULL)

Arguments

nsim	the number of simulations to run
scenario	all the survey information, result of build_sim
pred_dat	prediction data.frame
stratification	the stratification scheme to use for the stratified Horvtiz-Thompson estimate, specified as internal cutpoints
logit_opts	list of length 2, of parameters (scale and location) for the logistic that controls the behaviour of covariates when they are used. See build_sim_covar for more details.
transect_id	vector of the same length as the number of segments, used to group them for the Horvitz-Thompson estimate

Value

data.frame with one row per (successful) simulation. The data.frame has the following columns:

model

which model was fitted to the data

iter

which simulation this result is for

quantile

the quantile of the estimated distribution of abundance that the true abundance is in

Nhat

the estimated abundance

cvN

the coefficient of variation of the estimated abundance

n

number of observed objects

sp1

smoothing parameter 1

sp2

smoothing parameter 2

Author(s)

David L Miller

dill/ltdesigntester documentation built on May 15, 2019, 8:30 a.m.