Given a cpm object, calculate convenient descriptive statistics,
including the median CP, specified rI statistics, and pda
and pdb statistics for the fitted model (EoA parameterization), and
location and scale parameters for the fitted model (survival package
parameterization) along with estimated CIs.
desc(model_CP, Ir = c(1, 3, 7, 14, 28), CL = 0.9, nsim = 10000)A fitted CP model (cpm object)
The intervals for which to calculate the r statistics
The confidence level for the CIs.
Number of simulation draws for estimating CIs
Matrix of point and interval estimates for the median CP and the r
statistics for the specified intervals. The matrix is assigned to class
descCP that is simply a matrix with dimensions
ncell x (1 + 3*(5 + length(Ir))), column names that give the number of
observations in each cell, statistic name and upper and lower bounds
(in triplets), and row names giving the names of the cells. CL, nsim,
and the name of the fitted model (model_CP) are included as object
attributes.
The CIs for the r statistics (and the medianCP for the Weibull) ara
based on simulation of the pda and pdb parameters, calculation
of the statistics, and taking the empirical distribution of the simulated
values. Other CIs are based on the assumed bivariate normal distributions of
the appropriately transformed l and s parameters in the fitted
model using beta_hat and varbeta.
NOTE: rI is the probability that a carcass that arrives at a uniform random
time in an interval of I days will persist until the first search after
arrival.