Fit an SNR-detection function with a closed population capture-recapture glm

VGLM SNR-detection functions are positive bernoulli models of the form (y1,y2,...yN) ~ SNR. Here yi is a column of detections from the ith observer (with a 1 for detected and 0 for not detected by that observer). The assumptions for these models is that detection probability depends on heterogeneity from SNR, observers are independent, and there are no false positive detections included in the data for either observer. Models are fitted using the VGAM package (Yee, Stoklosa & Huggins 2015).

Yee, Thomas W., Jakub Stoklosa, and Richard M. Huggins. “The VGAM Package for Capture-Recapture Data Using the Conditional Likelihood.” Journal of Statistical Software 65 (June 1, 2015): 1–33. https://doi.org/10.18637/jss.v065.i05.

Usage

fitSNRvglm(
  SNRinfo,
  yColNames = c("detect_observer1", "detect_observer2"),
  whichObserver = "detect_observer2"
)

Arguments

SNRinfo

A data.frame containing detection information in columns Detected and SNR

yColNames

a list of strings containing the column names for each set of observer detections (default: 'detect_observer1','detect_observer2')

whichObserver

column name of the observer to use for predictions