The original ATLAS program (http://www.cbr.washington.edu/analysis/apps/atlas) provides the ability to analyze mark-recapture data that used active-tag technology, and to adjust the observed survival estimates for potential tag-failure that may be misidentified as a mortality of the marked animal. The cbrATLAS R package was created to allow this analysis be conducted in R. In addition, cbrATLAS incorporates an update of the original version to include the R package failCompare, increasing the number of model fitting options of tag-life data to predict tag-life of the active tags used in the study. The failCompare package also provides goodness-of-fits and rankings of the various models, to assist selection of the most appropriate model of tag-life, along with plotting the expected tag-life curve against observed tag failures.
cbrATLAS is currently capable of analyzing single release data histories (with and without censoring events) that may or may not require correction for active-tag failure. Future iterations are planned to incorporate paired release , Virtual-Paired Release (ViPRe), Virtual Release Dead Fish Correction (ViRDCt) study designs.
Please see the vignette for examples of a single release mark-recapture analysis, with and without correction for potential tag failure.