Roll back changes to GDR lambda calculations from 7c65a32f8, they result in bad simulated datasets

3 files changed, 34 insertions, 11 deletions

diff --git a/DESCRIPTION b/DESCRIPTION
index 5423c29..1bbd46b 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: unmarked
-Version: 1.2.5.9005
-Date: 2022-08-22
+Version: 1.2.5.9006
+Date: 2022-08-24
 Type: Package
 Title: Models for Data from Unmarked Animals
 Authors@R: c(
diff --git a/R/gdistremoval.R b/R/gdistremoval.R
index 43295f4..dc91934 100644
--- a/R/gdistremoval.R
+++ b/R/gdistremoval.R
@@ -517,9 +517,10 @@ setMethod("fitted", "unmarkedFitGDR", function(object){
   T <- object@data@numPrimary
 
   # Adjust log lambda when there is a random intercept
-  loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
-  loglam <- E_loglam(loglam, object, "lambda")
-  lam <- exp(loglam)
+  #loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
+  #loglam <- E_loglam(loglam, object, "lambda")
+  #lam <- exp(loglam)
+  lam <- predict(object, "lambda", level=NULL)$Predicted
   if(object@output == "density"){
     ua <- getUA(object@data)
     A <- rowSums(ua$a)
@@ -587,9 +588,10 @@ setMethod("ranef", "unmarkedFitGDR", function(object){
 
   Kmin = apply(ysum, 1, max, na.rm=T)
 
-  loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
-  loglam <- E_loglam(loglam, object, "lambda")
-  lam <- exp(loglam)
+  #loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
+  #loglam <- E_loglam(loglam, object, "lambda")
+  #lam <- exp(loglam)
+  lam <- predict(object, "lambda", level=NULL)$Predicted
   if(object@output == "density"){
     ua <- getUA(object@data)
     A <- rowSums(ua$a)
@@ -644,9 +646,10 @@ setMethod("ranef", "unmarkedFitGDR", function(object){
 setMethod("simulate", "unmarkedFitGDR", function(object, nsim, seed=NULL, na.rm=FALSE){
 
   # Adjust log lambda when there is a random intercept
-  loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
-  loglam <- E_loglam(loglam, object, "lambda")
-  lam <- exp(loglam)
+  #loglam <- log(predict(object, "lambda", level=NULL)$Predicted)
+  #loglam <- E_loglam(loglam, object, "lambda")
+  #lam <- exp(loglam)
+  lam <- predict(object, "lambda", level=NULL)$Predicted
   if(object@output == "density"){
     ua <- getUA(object@data)
     A <- rowSums(ua$a)
diff --git a/R/simulate.R b/R/simulate.R
index 5df7247..a8887cb 100644
--- a/R/simulate.R
+++ b/R/simulate.R
@@ -64,6 +64,7 @@ setMethod("simulate", "character",
   model <- blank_umFit(object)
   fit <- suppressWarnings(simulate_fit(model, formulas, guide, design, ...))
   coefs <- check_coefs(coefs, fit)
+  #fit <- replace_sigma(coefs, fit)
   coefs <- generate_random_effects(coefs, fit)
   fit <- replace_estimates(fit, coefs)
   ysims <- suppressWarnings(simulate(fit, nsim))
@@ -84,6 +85,25 @@ setMethod("simulate", "character",
   umfs
 })
 
+# Insert specified random effects SD into proper S4 slot in model object
+# This is mostly needed by GDR which uses the SD to calculate
+# N with E_loglam (this is currently disabled so the function is not needed)
+#replace_sigma <- function(coefs, fit){
+#  required_subs <- names(fit@estimates@estimates)
+#  formulas <- sapply(names(fit), function(x) get_formula(fit, x))
+#  rand <- lapply(formulas, lme4::findbars)
+#  if(!all(sapply(rand, is.null))){
+#    rvar <- lapply(rand, function(x) unlist(lapply(x, all.vars)))
+#    for (i in required_subs){
+#      if(!is.null(rand[[i]][[1]])){
+#        signame <- rvar[[i]]
+#        old_coefs <- coefs[[i]]
+#        fit@estimates@estimates[[i]]@randomVarInfo$estimates <- coefs[[i]][[signame]]
+#      }
+#    }
+#  }
+#  fit
+#}
 
 generate_random_effects <- function(coefs, fit){
   required_subs <- names(fit@estimates@estimates)