Unfortunately, we currently cannot guarantee that our implementation of the FBD-range process with (dnFBDSP) or without (dnFBDRP) tree inference yields correct likelihoods – in fact, we are quite sure that the former doesn't, and while the latter may well be correct or only require minor fixes, it hasn't been thoroughly validated. Therefore, this PR:
Removes the two tutorials we currently have on this topic ("Combined-Evidence Analysis and the Fossilized Birth-Death Process for Stratigraphic Range Data" and "Macroevolutionary Analysis of Stratigraphic Range Data" for the tree-inference and tree-free versions, respectively). Note that a third tutorial employing dnFBDRP ("Estimating speciation times using the fossilized birth-death range process") is already unindexed.
Removes the references to these tutorials, or any statements suggesting that the FBD-range process is implemented in RevBayes, from the remaining tutorials. Of these, one is indexed ("Combined-Evidence Analysis and the Fossilized Birth-Death Process for Analysis of Extant Taxa and Fossil Specimens") and two are unindexed ("Estimating a Time-Calibrated Phylogeny of Fossil and Extant Taxa using Morphological Data" and "Estimating speciation times using node dating").
Changes some wording in "Combined-Evidence Analysis and the Fossilized Birth-Death Process for Analysis of Extant Taxa and Fossil Specimens" to make it a bit clearer when to use specimen-level FBD and when to use FBD-range. The current version basically says "your age range can represent uncertainty or a stratigraphic range", and then proceeds to explain two paragraphs later why using stratigraphic ranges is actually a bad idea. I thought it better to make this clear from the beginning.
dwbapst
commented
3 weeks ago
Unfortunately, we currently cannot guarantee that our implementation of the FBD-range process with (
dnFBDSP) or without (dnFBDRP) tree inference yields correct likelihoods – in fact, we are quite sure that the former doesn't, and while the latter may well be correct or only require minor fixes, it hasn't been thoroughly validated. Therefore, this PR:Removes the two tutorials we currently have on this topic ("Combined-Evidence Analysis and the Fossilized Birth-Death Process for Stratigraphic Range Data" and "Macroevolutionary Analysis of Stratigraphic Range Data" for the tree-inference and tree-free versions, respectively). Note that a third tutorial employing
dnFBDRP("Estimating speciation times using the fossilized birth-death range process") is already unindexed.Removes the references to these tutorials, or any statements suggesting that the FBD-range process is implemented in RevBayes, from the remaining tutorials. Of these, one is indexed ("Combined-Evidence Analysis and the Fossilized Birth-Death Process for Analysis of Extant Taxa and Fossil Specimens") and two are unindexed ("Estimating a Time-Calibrated Phylogeny of Fossil and Extant Taxa using Morphological Data" and "Estimating speciation times using node dating").
Changes some wording in "Combined-Evidence Analysis and the Fossilized Birth-Death Process for Analysis of Extant Taxa and Fossil Specimens" to make it a bit clearer when to use specimen-level FBD and when to use FBD-range. The current version basically says "your age range can represent uncertainty or a stratigraphic range", and then proceeds to explain two paragraphs later why using stratigraphic ranges is actually a bad idea. I thought it better to make this clear from the beginning.