Changes to FSA — Estimating Abundance

I mentioned previously, that I have been updating the Mark-Recapture vignettes.  That has morphed into a document that is an update of the Mark-Recapture Closed and Open vignettes and the depletion/removal vignettes and associated FSA functions.  Some of the changes are cosmetic, others are more important.  I attempt to highlight the changes below but you should also see the NEWS file for the FSA package.  The new vignette is here.

FSA changes

• capHistConvert() — major changes, should consider this to be a new function.
• capHistSum() — added an m.array object to the output, changed column and row labels to the methodB objects, and added a plot() object to make the diagnostic plots identified in the Rcapture package.  Old code should still work.
• depletion() — Changed type= to method= and added DeLury as an option to method= (and left Delury). Changed ricker.mod= to Ricker.mod=.  Old code that explicitly used type= will be broken (need to change to method=).
• mrClosed() — Changed incl.inputs= to verbose=, changed type= to method=.  Added the ability to construct a CI for the overall PE when multiple groups are used in a Petersen family method (thus, added a incl.all= to confint()). Changed default for incl.all= from FALSE to TRUE.  Old code that explicitly used type= will be broken (need to change to method=).
• mrOpen() — Changed ci.type= to type= and phi.type= to phi.full=.  Removed type= from summary() and added a verbose= which will print only the estimates if FALSE or both observables and estimates if TRUE.  Added a verbose= to confint() to control whether the message about the type of confidence interval is printed or not.  Added another function, jolly(), that is the same as mrOpen().  Old code should work unless ci.type= and phi.type= were used in mrOpen() or type= was used in summary() .
• removal() — Added options to perform Moran (1951) and Schnute (1983) removal methods.  Completely modified the code so that the examples with apply() and lapply() would also provide confidence intervals. Changed so that if catch is a one row or one column matrix or data.frame then it will be converted to a vector, which allows one to extract one row from a data.frame to send to removal() without having to use as.numeric().  Old code should still work.

New Abundance Vignette

Again, this is now a combination of the original M-R closed, M-R open, and depletion/removal events.  All sections are like the original except for the changes in FSA from above and the theoretical development is greatly reduced.

The M-R closed section now includes an illustration to estimate N using the Petersen methods if the data is grouped (i.e., separate estimates by length groups) and for fitting the Otis et al. (1978) log-linear models to multiple M-R data with possible capture probability heterogeneities (using the Rcapture package).

The M-R open section now includes a section on fitting log-linear models (using Rcapture package).

The depletion/removal now includes is a section on using the method of Schnute (1983) to identify a statistically different capture probability for the first sample.