Generate a table of frequencies of probability of detection from a detection function model. This is particularly useful when employing covariates, as it can indicate if there are detections with very small detection probabilities that can be unduly influential when calculating abundance estimates.
Value
a data.frame with probability bins, counts and (optionally)
proportions. The object has an attribute p_range which contains the
range of estimated detection probabilities
Details
Because dht uses a Horvitz-Thompson-like estimator, abundance
estimates can be sensitive to errors in the estimated probabilities. The
estimator is based on \(\sum 1/ \hat{P}_a(z_i)\), which means that the
sensitivity is greater for smaller detection probabilities. As a rough
guide, we recommend that the method be not used if more than say 5% of the
\(\hat{P}_a(z_i)\) are less than 0.2, or if any are less than 0.1. If
these conditions are violated, the truncation distance w can be reduced.
This causes some loss of precision relative to standard distance sampling
without covariates.
References
Marques, F.F.C. and S.T. Buckland. 2004. Covariate models for the detection function. In: Advanced Distance Sampling, eds. S.T. Buckland, D.R. Anderson, K.P. Burnham, J.L. Laake, D.L. Borchers, and L. Thomas. Oxford University Press.
Examples
if (FALSE) { # \dontrun{
# try out the tee data
data(book.tee.data)
egdata <- book.tee.data$book.tee.dataframe
# fit model with covariates
result <- ddf(dsmodel = ~mcds(key = "hn", formula = ~sex+size),
data = egdata[egdata$observer==1, ], method = "ds",
meta.data = list(width = 4))
# print table
p.dist.table(result)
# with proportions
p.dist.table(result, proportion=TRUE)
} # }