2018-05-01 George G. Vega Yon
* R/pies.r: All dependencies have been moved to polygons.
* R/predict.r: idem.
2018-04-17 George G. Vega Yon
* src/phylo.cpp: Removed prediction functions (wrongly implemented).
* R/predict-bis.r: Documenting posterior probabilities.
* R/predict.r: Updating predict methods to use the correct version of
the posterior probabilities.
* tests/testthat/test-predict.r: Adding new tests.
2018-04-11 George G. Vega Yon
* src/predict.cpp: New function implementing posterior probabilities.
Now we are using pre-order to compute such probs (see paper).
* R/predict-bis.r: Idem. Including a brute-force version intended
to be used for testing.
* src/*: Fixing iter++ to ++iter when necesary (an bad practice that
now warns when compiling).
2018-02-28 George G. Vega Yon
* tests/*: Adding test checking that likelihood in model with eta =
1/2 is proportional to the likelihood without etas.
2018-02-27 George G. Vega Yon
* src/phylo.cpp: Updating etas in the model. Now it is the probability
of not reporting a discovery given the observed state (before it was
the true state).
* src/sim_phylo.cpp: idem.
* tests/*: idem.
2018-02-22 George G. Vega Yon
* src/phylo.cpp: Fixing 'error' back.
* R/predict.r: Adding eta to the call.
* R/plot_loglike.r: idem.
* R/estimation.r: idem, and adding function to try to compute inv.
* tests/testthat/*.r: Adding eta to the tests.
2018-02-21 George G. Vega Yon
* src/phylo.cpp: Adding the eta parameter for controlling for
annotation bias.
* src/sim_phylo.cpp: Idem.
* R/estimation.r: Idem.
* R/simulation.r: Idem.
* R/plot_loglike.r: Idem.
* R/loglike.r: Idem.
2018-01-25 George G. Vega Yon
* R/simulation.r: Values of `informative` and `maxtries` were not
passed
to simulate_fun_on_tree.
2018-01-18 George G. Vega Yon
* R/estimation.r: Fixing checking of non-informative data.
* R/predict.r: Indexing of edges was wrong.
* src/phylo.cpp: predict_fun was taking the wrong id for root node.
Also, when the node was the rootnode, d0i should yield probabilities
equal to pi.
* R/data-management.r: Removing functions that are no longer required.
* src/tree.cpp: Idem.
* tests/testthat/test-data-management.r: New set of tests.
2018-01-11 George G. Vega Yon
* R/ape.r: Fixing case in which as.phylo.matrix processes a tree with
only leafs and a root.
* R/data-management.r: Fixing bug in plot.aphylo.
2018-01-10 George G. Vega Yon
* R/estimation.r: Doc + estimation checks whether aphylo is informative.
* R/prune.r: Completely removed. No longer needed.
* R/simulation.r: Working on Doc + sim functions check if the resulting
tree is informative or not + new function rdrop_annotations.
* tests/testthat/*: Several new tests on the previous point.
2018-01-09 George G. Vega Yon
* R/ape.r: Working on documentation.
* R/data-management.r: idem
* R/simulation.r: idem.
* R/predict.r: Adapting pred functions to new structure of aphylo.
2018-01-04 George G. Vega Yon
* src/misc.cpp: New function reduce_pseq proposes a peeling seq that
skips ininformative nodes.
* R/data-management.r: Remove a bunch of routines that wont be used
anymore related to PO trees.
2018-01-03 George G. Vega Yon
* src/sim_phylo.cpp: Simulating trees using ape's phylo convention.
about 20x faster than rtree. Returns a phylo object.
* R/simulation.r: Idem.
* src/tree.cpp: Function to list offspring of a tree.
* R/data-management.r: idem.
* src/phylo.cpp: Likelihood was not been computed at the root node!
but this is for the new method. The old one was OK.
*
2017-12-20 George G. Vega Yon
* playground/*: Working on functions to prune trees using po traversal
or pre-order traversal. Algorithms working and good exercise to have
a better understanding of the pruning algorithm.
2017-10-18 George G. Vega Yon
* R/data-management.r: Fixing bug in plot.aphylo.
* playground/presentations: Working on EAC presentation.
2017-10-04 George G. Vega Yon
* R/data-management.r: Working in a standard method to point to
nodes ids.
* R/predict.r: Idem.
* R/prune.r: New example prunning leafs.
2017-10-03 George G. Vega Yon
* R/data-management.r: New constructor method for class po_tree.
Now it also includes branch lengths.
* R/simulation.r: Move documentation of sim_tree.
* src/sim_tree.cpp: Efficiency.
* src/tree.cpp: Change sorting in final edgelist.
2017-09-28 George G. Vega Yon
* R/data-management.r: Removing names for the po_tree labels.
* R/prune.r: New (a bit complex) function to prune trees.
2017-09-07 George G. Vega Yon
* R/data-management.r: leafs now has its own manual, plus
new internal checks checking labels.
* src/phylo.cpp: -noffspring- removed since now we use Rf_length.
This included changes in all the functions that used that argument.
* R/simulation.r: po_tree now has the labels as attributes, including
the offspring.
* src/sim_phylo.cpp: sim_tree now returns an integer matrix (edgelist).
Only offspring and labels are included as attributes.
2017-09-07 George G. Vega Yon
* R/panther.r: Now the user can choose which one to use.
* playground/ape_now_supports_singletons*: A small benchmark between
rncl and ape.
2017-09-05 George G. Vega Yon
* R/panther.r: Now using nrcl to read newick
2017-08-31 George G. Vega Yon
* R/pies.r: New function to rescale y-coords so that aspect ratio
is kept.
2017-08-29 George G. Vega Yon
* R/panther.r: Capturing info from newick extended version.
* R/pies.r: Adding more features to the pieplot.
* R/predict.r: New plot method for prediction.
* configure.ac: Compaling using OpenMP.
* condigure: idem.
* src/Makevars.in: idem.
2017-08-04 George G. Vega Yon
* R/panther.r: New function to read panther trees' annotation.
2017-08-01 George G. Vega Yon
* R/data-management.r: Working on aphylo-class. Plot method
using ggtree.
* R/estimation.r: Removing ABC method and cleaning code.
* tests/*: Adding more tests.
2017-07-31 George G. Vega Yon
* Overall, working on the -aphylo- class object. Now it is better
integrated with ape-phylo objects. Also, including class checks
for the class creation method.
2017-07-12 George G. Vega Yon
* playground/simulations: Moved to a different project aphylo_paper.
* src/phylo.cpp: Changed Pi to be -double- no longer an -arma::mat-.
* R/mle.r, R/plog_loglike.r: Idem.
* tests/testthat/test-loglikelihood.r: Idem.
2017-07-11 George G. Vega Yon
* src/quality.cpp, R/predict.r: Included new function to compute
expected value of random imputation given parameter alpha.
2017-05-31 George G. Vega Yon
* R/simulations.r: Fixed a small bug related to names of a vector.
2017-05-02 George G. Vega Yon
* R/mle.r: Added new functions to predict and evaluate prediction.
Plus, some code improvement.
* R/plot_loglike.r: Making code more clear.
* src/phylo.cpp: Prediction functions.
* src/misc.cpp: Function to compute geodesics.
* vignettes/*: Working on doc, including data imputation.
2017-04-19 George G. Vega Yon
* R/mle.r: Using fnscale and parscale in optim and abc_optim.
Now the functions maximize so its all out of the box.
2017-04-18 George G. Vega Yon
* *: Moved the MCMC to a new project: USCbiostats/amcmc
2017-04-13 George G. Vega Yon
* R/data-management.r: Functions as.phylo and recode_vertices
rewritten and renamed to as.apephylo and as_ape_tree. Also
now all use an edgelist instead of offspring and parent vectors.
* src/phylo.cpp: Changing name from Z to annotations (makes more sense).
* R/mle.r: Idem.
* src/sim_phylo.cpp: Name conventions, again, now edgelists are always
in the form of (parent, offspring).
* src/tree.cpp: Function to relabel edgelist so that these form a
partial ordered tree. Plus more doc.
2017-04-12 George G. Vega Yon
* src/tree: New functions for recoding trees as PO.
* R/*: Changing language... now phylo_tree is aphylo (Annotated
phylogenetic tree).
* R/data-management.r: get_offspring changed to new_aphylo, and
now includes more checks on the edge and annotations.
2017-03-29 George G. Vega Yon
* R/mcmc.r: Moved phylo_mcmc to R/mle.r
* R/mle.r: MLE now uses stats::optim rather than own NR.
* tests/testthat/*: new tests
2017-03-22 George G. Vega Yon
* R/mcmc.r: Multiple parallel chains and passing extra arguments to
fun.
* R/mle.r: Working on printing method.
* R/simulation.r: Adding templates for doc.
* tests/*: Adding more tests.
* src/phylo.cpp: Cleaning code and collapsing prob loops.
2017-03-08 George G. Vega Yon
* R/mle.r: New print method + covar matrix in phylo_mle.
2017-03-07 George G. Vega Yon
* R/data-management.r: Working with -as.phylo- method.
* src/sim_phylo.cpp: Moving the sim_tree function there.
* src/sim_tree.cpp: . .
* playground/simulating_and_estimating_data.r: Check the performance
of the mle estimator.
* R/simulation.r: Implements the function for sim annotated phylo
trees.
* tests/testthat/test-data-management.r: New set of tests
2017-02-28 George G. Vega Yon
* src/sim_tree.cpp: A function to simulate Directed Acyclic Graphs
is faster than ape::rtree and more convenient since it creates
the labelling that we require.
2017-02-23 George G. Vega Yon
* R/mcmc.r: -scale- now a vector. New example using MVN dist.
* src/mcmc.cpp: idem.
* src/sim_phylo.cpp: Multiple functions can be generated at the time.
* vignettes/*: Just doing some cleaning.
2017-02-22 George G. Vega Yon
* R/data-management.r: -as.phylo- method for -phylo_offspring-,
and removed igraph links (replaced by -ape-).
* playground/FormatTreeR_24Oct2012.pdf: Formally describes
the phylo objects.
* src/sim_phylo.cpp: Function to simulate phylo trees.
2017-02-16 George G. Vega Yon
* R/mcmc.r: Fixing small bug in output.
* R/mle.r: Adding an example with MCMC and improving doc.
* playground/*.r: Some tests of the performance of MCMC
including speed comparisons against the mcmc::metrop
function.
2017-02-15 George G. Vega Yon
* src/mcmc_misc.cpp -> src/mcmc.cpp: New version of the mcmc
algorithm implemented using Rcpp.
* R/mcmc.r: New option -useCpp- allows using a C++ version
of the algorithm. Speed improvement of ~ 50%.
* R/mle.r: Working on documentation for the phylo_mcmc
function.
* playground/mcmc_benchmark...: Measuring speed gains
of new MCMC implementation.
2017-02-14 George G. Vega Yon
* R/mcmc.r: mcmc now is MCMC and returns an object of class mcmc
from the coda R package. Also, new argument -thin- allows
'thinning' the samples to reduce autocorrelation between
draws. Also, working on aperiodicity of MCMC.
2017-02-09 George G. Vega Yon
* src/mcmc_misc.cpp: Started to write Gibbs Sampler.
2017-02-03 George G. Vega Yon
* R/mcmc.r: Adding example and fixing travis (doc missing). Also
including check of -ub- and -lb- parameters.
* tests/testthat/test-mcmc.r: New test for MCMC.
2017-02-02 George G. Vega Yon
* R/mle.r: Improving doc. Now optim pars are passed via `control`.
mcmc using own method.
* src/mcmc_misc.cpp, R/mcmc.r: New function for implemting MCMC
with hastings ratio.
2017-02-01 George G. Vega Yon
* R/mle.r: Adding more examples (filling zeros), plus more tunning
options for the NR method.
2017-01-18 George G. Vega Yon
* R/mle.r, R/plot_loglike.r: Changing coding style.
* R/mle.r: Working on NR (now gives more reasonable ans
and identical to ABCoptim).
* src/phylo.cpp: leaf_prob can be computed faster now,
and gain_loss_prob changed to prob_mat.
* src/phylo.h: Adding header.
* tests/testthat/*.r: leaf_prob now uses mu instead of M.
2017-01-18 George G. Vega Yon
* R/mle.r: Improving doc + output (class) structure.
* R/phylogenetic.r: Adding more imports + doc data.
* test/testthat/*: Adding tests (making sure the ll works).
* readme.rmd: Adding code coverage, travis and appveyor.
* src/peeling_cpp.cpp: Improving documentation.
2017-01-11 George G. Vega Yon
* R/mle.r: Maximum Likelihood estimation function and methods.
* R/phylogenetic.r: Adding package requirements for MLE.
2017-01-11 George G. Vega Yon
* R/data-management.r: get_offspring now returns neat experiment
* R/plot_loglike.r: Fixed bug
* playground/example.r: Accomodating script to changes in
get_offspring
* playground/root_node_prob.rmd: Notes on root node.
2017-01-10 George G. Vega Yon
* R/data-management.r: Changing offsprings to offspring
* R/methods.r: Idem.
* R/phylogenetic.r: Adding new imports.
* playground/example.r: Fixing N-R optim.
* vignettes/phylogenetic_trees.rmd: Working on doc.
* src/peeling_rcpp.cpp: Adding beta dist prior for psi, and,
fixing bug on leaf probabilities.
* R/plot_loglike.r: New function to visualize LogLike.
2017-01-03 George G. Vega Yon
* R/data-management.r: Now experimental data is filled. and
adding new set of functions.
* vignettes/phylogemetic_trees.rmd: Working on doc.
2017-01-03 George G. Vega Yon
* data/*: Adding data more neatly.
* R/*.r: Adding new methods for printing and preparing data.
* vignettes/*: Using html_vignette output.
2016-12-20 George G. Vega Yon
* Initial commit.