For the Beta Binomial distribution, we navigate to the section in the file with the header ‘Distributions’ and then look for the sub-header dealing with ‘math distributions’. Then, add the following line of code:
-#include "Dist_betaBinomial.h"
+ #include "Dist_betaBinomial.h"
This step registers the header file for the beta binomial distribution, effectively adding it to RevBayes.
diff --git a/developer/implementation/implement_functions.html b/developer/implementation/implement_functions.html
index 4e72f9a37..dd47d3aa1 100644
--- a/developer/implementation/implement_functions.html
+++ b/developer/implementation/implement_functions.html
@@ -67,9 +67,9 @@ Last modified on August 15, 2018
#ifndef HyperbolicCosineFunction_h
#define HyperbolicCosineFunction_h
-#include "ContinuousFunction.h"
-#include "TypedDagNode.h"
-#include <cmath>
+#include "ContinuousFunction.h"
+#include "TypedDagNode.h"
+#include <cmath>
namespace RevBayesCore {
/**
@@ -106,7 +106,7 @@ Last modified on August 15, 2018
Next, after including the necessary header files, we have to ensure that our new function is included within the RevBayesCore namespace.
Here we are implementing our hyperbolic cosine function as its own class that is derived from the continuous function class that is derived from the typed function class. This class stores the hyperbolic cosine of a value that is held in a DAG node. We have also defined a clone method which can create a clone of our class, and an update method which will update the value of our Hyperbolic Cosine class whenever the value of the DAG node changes. Now we will move on to the HyperbolicCosineFunction.cpp file.
-#include "HyperbolicCosineFunction.h"
+#include "HyperbolicCosineFunction.h"
using namespace RevBayesCore;
@@ -152,10 +152,10 @@ Last modified on August 15, 2018
#ifndef Func_hyperbolicCosine_h
#define Func_hyperbolicCosine_h
-#include "Real.h"
-#include "RlTypedFunction.h"
+#include "Real.h"
+#include "RlTypedFunction.h"
-#include <string>
+#include <string>
namespace RevLanguage {
@@ -199,12 +199,12 @@ Last modified on August 15, 2018
And the Func_hyperbolicCosine.cpp should look like this:
-#include "Func_hyperbolicCosine.h"
-#include "HyperbolicCosineFunction.h"
-#include "Probability.h"
-#include "Real.h"
-#include "RlDeterministicNode.h"
-#include "TypedDagNode.h"
+#include "Func_hyperbolicCosine.h"
+#include "HyperbolicCosineFunction.h"
+#include "Probability.h"
+#include "Real.h"
+#include "RlDeterministicNode.h"
+#include "TypedDagNode.h"
using namespace RevLanguage;
diff --git a/developer/implementation/index.html b/developer/implementation/index.html
index 04fe6e493..5f7f26d11 100644
--- a/developer/implementation/index.html
+++ b/developer/implementation/index.html
@@ -77,9 +77,9 @@ Overview
#ifndef HyperbolicCosineFunction_h
#define HyperbolicCosineFunction_h
-#include "ContinuousFunction.h"
-#include "TypedDagNode.h"
-#include <cmath>
+#include "ContinuousFunction.h"
+#include "TypedDagNode.h"
+#include <cmath>
namespace RevBayesCore {
/**
@@ -116,7 +116,7 @@ Overview
Next, after including the necessary header files, we have to ensure that our new function is included within the RevBayesCore namespace.
Here we are implementing our hyperbolic cosine function as its own class that is derived from the continuous function class that is derived from the typed function class. This class stores the hyperbolic cosine of a value that is held in a DAG node. We have also defined a clone method which can create a clone of our class, and an update method which will update the value of our Hyperbolic Cosine class whenever the value of the DAG node changes. Now we will move on to the HyperbolicCosineFunction.cpp file.
-#include "HyperbolicCosineFunction.h"
+#include "HyperbolicCosineFunction.h"
using namespace RevBayesCore;
@@ -162,10 +162,10 @@ Overview
#ifndef Func_hyperbolicCosine_h
#define Func_hyperbolicCosine_h
-#include "Real.h"
-#include "RlTypedFunction.h"
+#include "Real.h"
+#include "RlTypedFunction.h"
-#include <string>
+#include <string>
namespace RevLanguage {
@@ -209,12 +209,12 @@ Overview
And the Func_hyperbolicCosine.cpp should look like this:
-#include "Func_hyperbolicCosine.h"
-#include "HyperbolicCosineFunction.h"
-#include "Probability.h"
-#include "Real.h"
-#include "RlDeterministicNode.h"
-#include "TypedDagNode.h"
+#include "Func_hyperbolicCosine.h"
+#include "HyperbolicCosineFunction.h"
+#include "Probability.h"
+#include "Real.h"
+#include "RlDeterministicNode.h"
+#include "TypedDagNode.h"
using namespace RevLanguage;
@@ -317,7 +317,7 @@ General info before getting started
All mathematical distributions that have been implemented exist in core/distributions/math.
- Note that when implementing a new distribution, you will need to create .cpp and .h files in both the revlanguage directory and the core directory. (For a refresher on the difference between these two directories, refer to the {% page_ref architecture %} section of this Developer’s guide).
+
Note that when implementing a new distribution, you will need to create .cpp and .h files in both the revlanguage directory and the core directory. (For a refresher on the difference between these two directories, refer to the Getting familiar with the code section of this Developer’s guide).
The overall naming format remains the same for every distribution in RevBayes. In the Beta Binomial Distribution example provided below, I specify what naming convention to use when creating each file.
@@ -332,7 +332,7 @@ General info before getting started
Within every new distribution, you will need to include some functions. For example, each new distribution must have: the get class type, name, and help functions. You may not need to implement these from scratch (if they’re dictated by the parent class & are already present), but you will need to implement other functions within your distribution (e.g. cdf, rv, quantile).
- Distributions have a prefix DN (dag node), and all moves have a prefix MV. RevBayes takes the name within & creates the DN automatically, so be aware of this. For a refresher on DAG nodes, refer to {% page_ref architecture %}.
+ Distributions have a prefix DN (dag node), and all moves have a prefix MV. RevBayes takes the name within & creates the DN automatically, so be aware of this. For a refresher on DAG nodes, refer to Getting familiar with the code.
@@ -376,7 +376,7 @@ Steps
For the Beta Binomial distribution, we navigate to the section in the file with the header ‘Distributions’ and then look for the sub-header dealing with ‘math distributions’. Then, add the following line of code:
- #include "Dist_betaBinomial.h"
+ #include "Dist_betaBinomial.h"
This step registers the header file for the beta binomial distribution, effectively adding it to RevBayes.
@@ -397,18 +397,18 @@ Steps
There are multiple ways to do this. As a best practice, you should first compile it to ensure there are no errors. Once it compiles with no problems, you can test in various ways (e.g. run each individual function within the new Beta Binomial distribution in R, then run the Binomial distribution with a Beta prior in Rev and see if the output matches). For more information, see the developer tutorials on validation and testing. (TODO)
- After ensuring your code runs properly, you are ready to add it to the git repo. We recommend reading through the {% page_ref git-flow %} section of the Developer’s guide before pushing.
+ After ensuring your code runs properly, you are ready to add it to the git repo. We recommend reading through the RevBayes Git Workflow section of the Developer’s guide before pushing.
Implementing a Metropolis-Hastings Move
-General info before getting started
+General info before getting started
The steps to implementing a new move vary slightly, depending on the move’s type (e.g., Metropolis-Hastings versus Gibbs). For the purpose of this guide, we will focus on a Metropolis-Hastings move.
In general, the fastest and easiest way to get help is to find the most similar move already implemented in RevBayes and use it as a guide. Remember that, as with implementing a new distribution or function, you’ll need to add relevant code to both the core of RevBayes and the language. Also remember that you’ll need to work out the math appropriate for your move (e.g., the Hastings ratio) ahead of time.
-Steps
+Steps
-
diff --git a/developer/setup/index.html b/developer/setup/index.html
index 4683d3d10..7a883138d 100644
--- a/developer/setup/index.html
+++ b/developer/setup/index.html
@@ -93,7 +93,7 @@
Getting RevBayes working
Setting up Eclipse for RevBayes development
Eclipse is a Java-based, cross-platform IDE with lots of nice features that make it convenient for RevBayes development. First, it’s cross-platform, and unlike Xcode, you can use any compiler you like (not just clang).
-Prerequisites
+Prerequisites
- To compile RevBayes using Eclipse, you must install CMake and its command line tools. This is very easy if you have homebrew installed (
brew install cmake)
- You must also have a C++ compiler. If you are using a Mac, you can use
clang (included with XCode), or you can install gcc (e.g. via homebrew brew install gcc).
@@ -264,7 +264,7 @@ Check C++ language options
Set up the XCode Project with an external build system
-Prerequisites
+Prerequisites
- To compile RevBayes using Xcode, first install https://cmake.org/ and its https://stackoverflow.com/questions/30668601/installing-cmake-command-line-tools-on-a-mac. This is easily done using https://brew.sh/ by running
brew install cmake in terminal.
diff --git a/tutorials/biogeo/scripts/make_anc_state.Rev b/tutorials/biogeo/scripts/make_anc_state.Rev
index 4121e3013..2c8fa1ac1 100644
--- a/tutorials/biogeo/scripts/make_anc_state.Rev
+++ b/tutorials/biogeo/scripts/make_anc_state.Rev
@@ -42,13 +42,12 @@ anc_states_stoch_map = readAncestralStateTrace(out_stoch_fn)
# make summary tree
-char_map_tree = characterMapTree(tree=mcc_tree,
- ancestral_state_trace_vector=anc_states_stoch_map,
- character_file=out_str+".char.map.tre",
- posterior_file=out_str+".pp.map.tre",
- burnin=0.25,
- num_time_slices=500)
-
+# char_map_tree = characterMapTree(tree=mcc_tree,
+# ancestral_state_trace_vector=anc_states_stoch_map,
+# character_file=out_str+".char.map.tre",
+# posterior_file=out_str+".pp.map.tre",
+# burnin=0.25,
+# num_time_slices=500)
# exit RevBayes
quit()
diff --git a/tutorials/model_selection_bayes_factors/bf_intro.html b/tutorials/model_selection_bayes_factors/bf_intro.html
index 39ffa5345..8d253e07d 100644
--- a/tutorials/model_selection_bayes_factors/bf_intro.html
+++ b/tutorials/model_selection_bayes_factors/bf_intro.html
@@ -55,7 +55,7 @@
General Introduction to Model selection
Comparing relative model fit with Bayes factors
Sebastian Höhna, Michael J Landis, Tracy A Heath and Brian R Moore
- Last modified on September 12, 2019
+ Last modified on April 5, 2024
@@ -328,7 +328,8 @@ Estimating the Margi
model here. Our main motivation is to show that the marginal likelihood
estimation algorithms are independent of the model. Thus, you can apply
these algorithms to any model, e.g., relaxed
-clock models and birth-death models, as well.
+clock models and birth-death models, as well.
+For a more concrete example including data and Rev script, please see the follow-up tutorial Model selection of common substitution models for one locus.
Computing Bayes Factors and Model Selection
diff --git a/tutorials/model_selection_bayes_factors/bf_partition_model.html b/tutorials/model_selection_bayes_factors/bf_partition_model.html
index d334d9a88..1ea2a14e4 100644
--- a/tutorials/model_selection_bayes_factors/bf_partition_model.html
+++ b/tutorials/model_selection_bayes_factors/bf_partition_model.html
@@ -55,7 +55,7 @@
Model selection of partition models
Comparing relative model fit with Bayes factors
Mike May and Sebastian Höhna
- Last modified on September 12, 2019
+ Last modified on April 9, 2024
@@ -103,14 +103,6 @@ Data files and scripts
fagus_rbcL.nex
- sim_locus1.nex
-
- sim_locus2.nex
-
- sim_locus3.nex
-
- sim_locus4.nex
-
@@ -129,8 +121,6 @@ Data files and scripts
marginal_likelihood_partition_5.Rev
- ml_Partition_model.Bodega.Rev
-
diff --git a/tutorials/model_selection_bayes_factors/bf_subst_model.html b/tutorials/model_selection_bayes_factors/bf_subst_model.html
index 4f9a22802..6ce5a69f6 100644
--- a/tutorials/model_selection_bayes_factors/bf_subst_model.html
+++ b/tutorials/model_selection_bayes_factors/bf_subst_model.html
@@ -55,7 +55,7 @@
Model selection of common substitution models for one locus
Comparing relative model fit with Bayes factors
Sebastian Höhna, Michael J Landis, Tracy A Heath
- Last modified on September 12, 2019
+ Last modified on April 9, 2024
@@ -86,6 +86,34 @@ Overview
+
+ Data files and scripts
+
+
+ Data Files
+
+
+
+
+ - primates_and_galeopterus_cytb.nex
+
+
+
+
+ Scripts
+
+
+
+
+ - marginal_likelihood_GTR_Gamma_inv.Rev
+
+ - marginal_likelihood_JC.Rev
+
+
+
+
+
+
Overview
diff --git a/tutorials/model_selection_bayes_factors/bf_subst_model_rj.html b/tutorials/model_selection_bayes_factors/bf_subst_model_rj.html
index e89aeabe4..e36ab1920 100644
--- a/tutorials/model_selection_bayes_factors/bf_subst_model_rj.html
+++ b/tutorials/model_selection_bayes_factors/bf_subst_model_rj.html
@@ -55,7 +55,7 @@
Model averaging of substitution models
Reversible-jump MCMC over substitution models
Mike May and Sebastian Höhna
- Last modified on September 12, 2019
+ Last modified on April 9, 2024
@@ -97,13 +97,7 @@ Data files and scripts
- fagus_ITS.nex
-
- fagus_matK.nex
-
- fagus_rbcL.nex
-
- primates_and_galeopterus_cox2.nex
+ primates_and_galeopterus_cytb.nex