As a follow-up to my post about major changes to FSA and that some of the “old” vignettes are now out-of-date, here is a brief summary of new material in the draft book chapters (linked to below) that replaced the vignettes.
- Age Comparisons – Major changes are the addition of two more tests of symmetry (McNemar’s and Evans-Hoenigs), changed the
nref.lab=(easier to remember which one represents the reference data) in
ageBias(), added a
plot()to plot the number of fish at each point, and the correction of a bug in
agePrecision()that occurred when two ages agreed for an age-0 fish. Minor changes included adding
what="n"to get the sample size for the age-agreement table, adding the ability to use multiple
plot()to allow control of the position of the sample size values on the plot, changing the default color of the confidence intervals in
plot()to black, changing
col.CIsig=, and changing the defaults of the agreement line in
plot(). Several other internal changes were made but should be transparent to the user. The vignette is built upon a more interesting (and completely real) data set. All results were tested against published sources or results from other validated softwares.
- Weight-Length – No major changes. The vignette also serves as an introduction to simple linear and dummy variable regression so the vignette is longer to accommodate the general introduction to these methods.
- Growth – Major changes include the following to
vbStarts(): forcing the use of a formula, changing
ages2use=, changing the Schnute method to use the ages in
ages2use=rather than being hard-wired to use the minimum and maximum observed age, allowing both the Schnute and Francis methods to use the minimum and maximum observed ages if
Schnuteparameterization to use t3 and L3 instead of t2 and L2 (allows for simplicity in explanation with Francis model). The
growthModelSim()had similar changes relative to the Schnute parameterization and the user is forced to use a formula. Changed
"vbGQ", changed all
"Gompertz", and added a
"Weisberg"parameterization to all functions. Add the
lrt()functions for performing extra sum-of-squares and likelihood ratio tests when comparing nested models (these are more useful when comparing several nested models than
lrtest()). Tested all output against verified sources. Greatly improved (in my opinion) the description for comparing two growth models. This chapter also serves as an introduction to non-linear model fitting and, thus, is longer than the original vignette.
- Mark-Recapture – I have not finished the changes here (will likely be a while), but the major change thusfar has been changing the
mrClosed()and major changes to
capHistConvert()(too many changes to detail here … should treat this as a new function if updating your code).