multimark 2.1.6 2023-03-09
-----------------
CHANGES

* fix error on r-devel as requested by CRAN


multimark 2.1.5 2022-10-18
-----------------
CHANGES

* fix compilation warnings on linux with -Wstrict-prototypes in CFLAGS


multimark 2.1.4 2022-03-07
-----------------
CHANGES

* fix error on Fedora as requested by CRAN


multimark 2.1.3 2021-09-13
-----------------
CHANGES

* removed implicit conversions in C code that triggered compilation warnings on macOS with -Wconversion -Wno-sign-conversion in CFLAGS

BUG FIXES

* MCMC algorithm for multimarkClosedSCR() incorrectly calculated detection probabilities for the M - n members of the ``superpopulation'' of encountered individuals whenever their latent activity centers were updated. The activity center proposal distributions were also reversed when calculating the Metropolis-Hastings acceptance ratio for these individuals. These relatively minor bugs should in most cases have little impact on MCMC results, but previous analyses using multimarkClosedSCR() should be revisited to confirm


multimark 2.1.2 2020-02-05
-----------------
CHANGES

* remove unnecessary dependencies on doParallel, foreach, and doRNG

* tweaked tests in order to pass MKL case at CRAN


multimark 2.1.1 2019-11-25
-----------------
CHANGES

* for SCR models, the suggested tolerance limit from sp::points2grid is now used when 'studyArea' coordinate intervals are not constant

* for SCR models, studyArea rows are now forced to be ordered by raster cell order (i.e. left to right, top to bottom)


multimark 2.1.0 2019-04-21
-----------------
CHANGES

* clusterSetRNGStream now used for reproducible parallel processing

* added error trap to processdataSCR if Enc.Mat includes non-zero encounters on inactive traps

* NAMESPACE now generated using roxygen2

BUG FIXES

* when using processdataSCR to generate a study area, the number of grid cells are now automatically adjusted to ensure cells are square


multimark 2.0.1 2017-06-30
-----------------
NEW FEATURES

* Added bobcatSCR example data for spatial capture-recapture analysis using multimarkClosedSCR

BUG FIXES

* Typo corrections in help files


multimark 2.0.0 2016-05-23
-----------------
NEW FEATURES

* Added spatial capture-recapture model of Royle et al. (2009) using a semi-complete data likelihood (King et al. 2016)

* Added tiger example data for spatial capture-recapture analysis using markClosedSCR

* Increased capacity for handling hundreds of sampling occasions


multimark 1.3.2 2015-12-18
-----------------
BUG FIXES

* Allow delta_B = 0 (i.e. 1 - (delta_1 + delta_2) = 0) in 'simdataCJS()' and 'simdataClosed();


multimark 1.3.1 2015-10-20
-----------------
CHANGES

* Updated citations for 'multimark'

BUG FIXES

* Typo corrections in help files


multimark 1.3.0 2015-07-27
-----------------
NEW FEATURES

* 'markClosed()' and 'markCJS()' wrappers facilitate ``conventional'' capture-recapture analysis with a single mark type; Bayesian multimodel inference using 'multimodelClosed()' and 'multimodelCJS()' can also be implemented with output from these ``conventional'' models

CHANGES

* 'mms' no longer an argument for 'multimodelClosed()' or 'multimodelCJS()'; these data are now returned with individual model output for additional error checking

* CJS latent variable 'z' now denoted as 'q'

BUG FIXES

* Fixed bug in 'multimarkClosed()' when all encounter histories are known


multimark 1.2.0 2015-06-23
-----------------
NEW FEATURES

* Beta prior for 'psi' parameter can now be user specified in 'multimarkCJS()' and 'multimarkClosed()' using 'a0psi' and 'b0psi' arguments

* burn-in and thinning interval can now be user specified in 'multimodelCJS()' and 'multimodelClosed()' using 'mburnin' and 'mthin' arguments

CHANGES

* Minor bug fixes in updates for 'psi', 'sigma2_zp', and 'sigma2_zphi' in 'multimarkCJS()'

* developed using R version 3.2.1 (2015-06-18); 'multimark' now requires R >=3.2.1 for compatibility with changes to base function 'nchar()'


multimark 1.1.4 2015-04-29
-----------------
NEW FEATURES

* parallel processing now used in 'multimodelCJS()' and 'multimodelClosed()' for >1 chains

CHANGES

* parallel processing now only used with >1 chains in 'multimarkCJS()' and 'multimarkClosed()'

* modified chain progress printing for 'multimarkCJS()', 'multimarkClosed()', 'multimodelCJS()' and 'multimodelClosed()'

* temporal trend (Time) predictors centered around zero

* updated package citation

* modification to documentation to avoid overflow on pdf

* developed using R version 3.2.0 (2015-04-16)


multimark 1.1.3 2015-04-16
-----------------
* Initial CRAN posting