Add TNT-style transient autoconstraint (default OFF, opt-in)

R/SearchControl.R

@@ -206,6 +206,22 @@
 #'   uncertain regions.  Constraints are cleared whenever a new best score
 #'   is found.  Only active when no user-supplied `constraint` is
 #'   present.  Default `FALSE`.
+#' @param transientAutoconst Logical; use TNT-style *transient*
+#'   auto-constraint?  Distinct from the rigid `consensusConstrain`: each
+#'   replicate builds a per-replicate constraint from the strict consensus
+#'   of the previous replicates' pool trees intersected with this
+#'   replicate's own Wagner start, enforces it only over the early pipeline
+#'   stages (initial TBR through the first sectorial pass), then *releases*
+#'   it before the ratchet, drift, later sectorial and final TBR stages.
+#'   The first replicate of each hit (replicate 0, or any replicate that
+#'   immediately follows a new best score) is left unconstrained.  Because
+#'   the constraint refreshes every replicate and is never permanently
+#'   locked, it can guide early search without entrenching a basin.  Only
+#'   active when neither a user `constraint` nor `consensusConstrain` is in
+#'   effect.  Default `FALSE`.
+#' @param transientAutoconstMinReps Integer; number of completed replicates
+#'   (with at least two best-score pool trees) required before
+#'   `transientAutoconst` engages.  Default `2`.
 #' @param wagnerBias Integer; criterion for biasing taxon addition order
 #'   during Wagner tree construction.  0 = random (default),
 #'   1 = Goloboff (2014) non-ambiguous-character priority,
@@ -342,6 +358,8 @@ SearchControl <- function(
     perturbStopFactor = 2L,
     adaptiveLevel = FALSE,
     consensusConstrain = FALSE,
+    transientAutoconst = FALSE,
+    transientAutoconstMinReps = 2L,
     # Taxon pruning-reinsertion (T-266)
     pruneReinsertCycles = 0L,
     pruneReinsertDrop = 0.10,
@@ -441,6 +459,8 @@ SearchControl <- function(
       perturbStopFactor = as.integer(perturbStopFactor),
       adaptiveLevel = as.logical(adaptiveLevel),
       consensusConstrain = as.logical(consensusConstrain),
+      transientAutoconst = as.logical(transientAutoconst),
+      transientAutoconstMinReps = as.integer(transientAutoconstMinReps),
       pruneReinsertCycles = as.integer(pruneReinsertCycles),
       pruneReinsertDrop = as.double(pruneReinsertDrop),
       pruneReinsertSelection = as.integer(pruneReinsertSelection),
@@ -488,7 +508,8 @@ print.SearchControl <- function(x, ...) {
                      "poolMaxSize", "poolSuboptimal"),
     "Stopping" = c("consensusStableReps", "perturbStopFactor",
                     "adaptiveLevel",
-                    "consensusConstrain", "adaptiveStart",
+                    "consensusConstrain", "transientAutoconst",
+                    "transientAutoconstMinReps", "adaptiveStart",
                     "enumTimeFraction")
   )
   cat("SearchControl object\n")

R/ts-driven-compat.R

@@ -73,6 +73,8 @@ ts_driven_search <- function(
     perturbStopFactor = 2L,
     adaptiveLevel = FALSE,
     consensusConstrain = FALSE,
+    transientAutoconst = FALSE,
+    transientAutoconstMinReps = 2L,
     nniPerturbCycles = 0L,
     nniPerturbFraction = 0.5,
     wagnerBias = 0L,
@@ -134,6 +136,8 @@ ts_driven_search <- function(
     perturbStopFactor = as.integer(perturbStopFactor),
     adaptiveLevel = as.logical(adaptiveLevel),
     consensusConstrain = as.logical(consensusConstrain),
+    transientAutoconst = as.logical(transientAutoconst),
+    transientAutoconstMinReps = as.integer(transientAutoconstMinReps),
     pruneReinsertCycles = as.integer(pruneReinsertCycles),
     pruneReinsertDrop = as.double(pruneReinsertDrop),
     pruneReinsertSelection = as.integer(pruneReinsertSelection),

inst/REFERENCES.bib

@@ -1,25 +1,3 @@
-@article{LapointeCucumel1997,
-  title = {The average consensus procedure: combination of weighted trees containing identical or overlapping sets of taxa},
-  author = {Lapointe, Fran{\c c}ois-Joseph and Cucumel, Guy},
-  year = {1997},
-  journal = {Systematic Biology},
-  volume = {46},
-  number = {2},
-  pages = {306--312},
-  doi = {10.1093/sysbio/46.2.306}
-}
-
-@article{Arias2004,
-  title = {Profile parsimony ({{PP}}): an analysis under implied weights ({{IW}})},
-  author = {Arias, J. Salvador and {Miranda-Esquivel}, Daniel Rafael},
-  year = {2004},
-  volume = {20},
-  pages = {56--63},
-  doi = {10.1111/j.1096-0031.2003.00001.x},
-  journal = {Cladistics},
-  number = {1}
-}
-
 @article{Aberer2013,
   title = {Pruning rogue taxa improves phylogenetic accuracy: an efficient algorithm and webservice},
   author = {Aberer, Andre J. and Krompass, Desmithconsnis and Stamatakis, Alexandros},
@@ -42,6 +20,17 @@ @article{Archie1989
   doi = {10.2307/2992286}
 }
 
+@article{Arias2004,
+  title = {Profile parsimony ({{PP}}): an analysis under implied weights ({{IW}})},
+  author = {Arias, J. Salvador and {Miranda-Esquivel}, Daniel Rafael},
+  year = {2004},
+  volume = {20},
+  pages = {56--63},
+  doi = {10.1111/j.1096-0031.2003.00001.x},
+  journal = {Cladistics},
+  number = {1}
+}
+
 @article{Bien2011,
   title = {Hierarchical Clustering with Prototypes via Minimax Linkage},
   author = {Bien, Jacob and Tibshirani, Robert},
@@ -93,6 +82,27 @@ @article{Carter1990
   doi = {10.1137/0403005}
 }
 
+@article{Chao1984,
+  author    = {Chao, Anne},
+  title     = {Nonparametric estimation of the number of classes in a population},
+  journal   = {Scandinavian Journal of Statistics},
+  year      = {1984},
+  volume    = {11},
+  number    = {4},
+  pages     = {265--270}
+}
+
+@article{Chao2012,
+  author    = {Chao, Anne and Jost, Lou},
+  title     = {Coverage-based rarefaction and extrapolation: standardizing samples by completeness rather than size},
+  journal   = {Ecology},
+  year      = {2012},
+  volume    = {93},
+  number    = {12},
+  pages     = {2533--2547},
+  doi       = {10.1890/11-1952.1}
+}
+
 @article{Congreve2016,
   author = {Congreve, Curtis R. and Lamsdell, James C.},
   doi = {10.1111/pala.12236},
@@ -357,6 +367,17 @@ @article{Gonzalez1985
   doi     = {10.1016/0304-3975(85)90224-5}
 }
 
+@article{Good1953,
+  author    = {Good, I. J.},
+  title     = {The population frequencies of species and the estimation of population parameters},
+  journal   = {Biometrika},
+  year      = {1953},
+  volume    = {40},
+  number    = {3/4},
+  pages     = {237--264},
+  doi       = {10.2307/2333344}
+}
+
 @article{Hartigan1979,
   title = {Algorithm {{AS}} 136: A {{{\emph{K}}}}-Means Clustering Algorithm},
   shorttitle = {Algorithm {{AS}} 136},
@@ -400,6 +421,17 @@ @article{Kumar2012
   number = {2}
 }
 
+@article{LapointeCucumel1997,
+  title = {The average consensus procedure: combination of weighted trees containing identical or overlapping sets of taxa},
+  author = {Lapointe, Fran{\c c}ois-Joseph and Cucumel, Guy},
+  year = {1997},
+  journal = {Systematic Biology},
+  volume = {46},
+  number = {2},
+  pages = {306--312},
+  doi = {10.1093/sysbio/46.2.306}
+}
+
 @article{Maddison1991,
   title = {Null models for the number of evolutionary steps in a character on a phylogenetic tree},
   author = {Maddison, Wayne P. and Slatkin, M},
@@ -474,6 +506,17 @@ @article{Muller2005
   number = {1}
 }
 
+@article{Nguyen2015,
+  title = {{IQ-TREE}: A Fast and Effective Stochastic Algorithm for Estimating Maximum-Likelihood Phylogenies},
+  author = {Nguyen, Lam-Tung and Schmidt, Heiko A. and von Haeseler, Arndt and Minh, Bui Quang},
+  year = {2015},
+  volume = {32},
+  pages = {268--274},
+  doi = {10.1093/molbev/msu300},
+  journal = {Molecular Biology and Evolution},
+  number = {1}
+}
+
 @article{Nixon1999,
   author = {Nixon, Kevin C.},
   doi = {10.1111/j.1096-0031.1999.tb00277.x},
@@ -705,6 +748,17 @@ @article{Sun2018
   doi = {10.1098/rspb.2018.1780}
 }
 
+@article{Thompson1933,
+  title = {On the Likelihood that One Unknown Probability Exceeds Another in View of the Evidence of Two Samples},
+  author = {Thompson, William R.},
+  year = {1933},
+  volume = {25},
+  pages = {285--294},
+  doi = {10.2307/2332286},
+  journal = {Biometrika},
+  number = {3/4}
+}
+
 @article{Vinther2008,
   author = {Vinther, Jakob and {Van Roy}, Peter and Briggs, Derek E. G.},
   doi = {10.1038/nature06474},
@@ -753,57 +807,3 @@ @article{Wilkinson2017
   number = {4}
 }
 
-@article{Nguyen2015,
-  title = {{IQ-TREE}: A Fast and Effective Stochastic Algorithm for Estimating Maximum-Likelihood Phylogenies},
-  author = {Nguyen, Lam-Tung and Schmidt, Heiko A. and von Haeseler, Arndt and Minh, Bui Quang},
-  year = {2015},
-  volume = {32},
-  pages = {268--274},
-  doi = {10.1093/molbev/msu300},
-  journal = {Molecular Biology and Evolution},
-  number = {1}
-}
-
-@article{Thompson1933,
-  title = {On the Likelihood that One Unknown Probability Exceeds Another in View of the Evidence of Two Samples},
-  author = {Thompson, William R.},
-  year = {1933},
-  volume = {25},
-  pages = {285--294},
-  doi = {10.2307/2332286},
-  journal = {Biometrika},
-  number = {3/4}
-}
-
-@article{Chao1984,
-  author    = {Chao, Anne},
-  title     = {Nonparametric estimation of the number of classes in a population},
-  journal   = {Scandinavian Journal of Statistics},
-  year      = {1984},
-  volume    = {11},
-  number    = {4},
-  pages     = {265--270}
-}
-
-@article{Good1953,
-  author    = {Good, I. J.},
-  title     = {The population frequencies of species and the estimation of population parameters},
-  journal   = {Biometrika},
-  year      = {1953},
-  volume    = {40},
-  number    = {3/4},
-  pages     = {237--264},
-  doi       = {10.2307/2333344}
-}
-
-@article{Chao2012,
-  author    = {Chao, Anne and Jost, Lou},
-  title     = {Coverage-based rarefaction and extrapolation: standardizing samples by completeness rather than size},
-  journal   = {Ecology},
-  year      = {2012},
-  volume    = {93},
-  number    = {12},
-  pages     = {2533--2547},
-  doi       = {10.1890/11-1952.1}
-}
-