expanded description

README.md

@@ -6,7 +6,9 @@ Examine paleoTS model selection performance with time series length
 
 ## Introduction
 
-Hohmann et al 2023 ([DOI: 10.1101/2023.12.18.572098](https://doi.org/10.1101/2023.12.18.572098)), supplementary code ([DOI: 10.5281/zenodo.10390267](https://doi.org/10.5281/zenodo.10390267)) found unexpected behavior in the model selection performance of the paleoTS package, version 0.5.3 ([CRAN.R-project.org/package=paleoTS](https://CRAN.R-project.org/package=paleoTS)). Here, this behavior is examined using the simulation tools implemented in the paleoTS package.
+Hohmann et al 2023 ([DOI: 10.1101/2023.12.18.572098](https://doi.org/10.1101/2023.12.18.572098)), supplementary code ([DOI: 10.5281/zenodo.10390267](https://doi.org/10.5281/zenodo.10390267)) found unexpected behavior in the model selection performance of the paleoTS package, version 0.5.3 ([CRAN.R-project.org/package=paleoTS](https://CRAN.R-project.org/package=paleoTS)) when including Ornstein-Uhlenbeck processes in the set of modes of evolution tested for. Here, this behavior is examined using the simulation tools implemented in the paleoTS package.
+
+The code simulates stasis and undirected random walks, and examines how AICc weights for different modes of evolution (stasis, (un)directed random walk) change with time series length.
 
 ## Authors
 
@@ -52,8 +54,10 @@ Results of the analysis are already stored in the repository under _figs/_. You
 * code : folder for code
   * test.paleots.R : R script examining paleoTS model selection with time series length
 * figs : folder for figures
-  * test_stasis.jpeg : AICc weights under stasis model with increasing time series length
-  * test_urw.jpeg : AICc weights under undirected random walk (URW) model with increasing time series length
+  * test_stasis_with_ou.jpeg : AICc weights under stasis model with increasing time series length, including OU in the tested modes
+  * test_stasis_without_ou.jpeg : AICc weights under stasis model with increasing time series length, not including OU in the tested modes
+  * test_urw_with_ou.jpeg : AICc weights under undirected random walk (URW) model with increasing time series length, including OU in the tested modes
+  * test_urw_without_ou.jpeg : AICc weights under undirected random walk (URW) model with increasing time series length, not including OU in the tested modes
 * renv : folder for `renv` package
 * .Rprofile : R session info
 * .gitignore : untracked files

code/test.paleots.R

@@ -20,9 +20,9 @@ set.seed(1) # set seed for reproducibility
 
 #### stasis with test for OU ####
 
-cat("Running stasis simulations \n")
+cat("Running stasis simulations 1/2 \n")
 for (j in 1:9){
-  print(j)
+  cat(paste0("time series length ", ns[j],"\n"))
   for (i in 1:100){
 
     test<-paleoTS::sim.Stasis(ns=ns[j],
@@ -61,9 +61,9 @@ mtext('green = stasis, yellow = OU, blue = URW, red = GRW',side=3)
 dev.off()
 
 #### URW with test for OU #### 
-cat("Running URW simulations \n")
+cat("Running URW simulations 1/2 \n")
 for (j in 1:9){
-  print(j)
+  cat(paste0("time series length ", ns[j],"\n"))
   for (i in 1:100){
 
     test = paleoTS::sim.GRW(ns = ns[j], # time series length
@@ -105,9 +105,9 @@ compile<-array(dim = c(3,100,9))
 
 #### Stasis without test for OU ####
 
-cat("Running stasis simulations \n")
+cat("Running stasis simulations  2/2 \n")
 for (j in 1:9){
-  print(j)
+  cat(paste0("time series length ", ns[j],"\n"))
   for (i in 1:100){
 
     test<-paleoTS::sim.Stasis(ns=ns[j],
@@ -131,7 +131,7 @@ compile.group<-data.frame(
   stringsAsFactors = FALSE
 )
 compile.group$y<-ordered(compile.group$y,levels=c('Stasis','URW','GRW'))
-jpeg("figs/test_stasis.jpeg")
+jpeg("figs/test_stasis_without_ou.jpeg")
 boxplot(x~y+z,data = compile.group,
         col=c('lightgreen','dodgerblue3','firebrick3'),
         xaxt = 'n',
@@ -148,9 +148,9 @@ dev.off()
 
 #### URW without test for OU ####
 
-cat("Running URW simulations \n")
+cat("Running URW simulations 2/2 \n")
 for (j in 1:9){
-  print(j)
+  cat(paste0("time series length ", ns[j],"\n"))
   for (i in 1:100){
 
     test = paleoTS::sim.GRW(ns = ns[j], # time series length
@@ -174,7 +174,7 @@ compile.group<-data.frame(
 )
 compile.group$y<-ordered(compile.group$y,levels=c('Stasis','URW','GRW'))
 
-jpeg("figs/test_urw.jpeg")
+jpeg("figs/test_urw_without_ou.jpeg")
 boxplot(x~y+z,data = compile.group,
         col=c('lightgreen','dodgerblue3','firebrick3'),
         xaxt = 'n',