- New zero-handling procedure
- If
shrink = TRUE
, bin frequencies are estimated by James-Stein-type shrinkage
- If
- New objective functions
- Add
asSLR
support toobjective.keepWADIST
by comparing with Euclidean distance objective.maxDSKL
maximizes the RDA of a PCoA of the SKL divergences
- Add
- New objective function
objective.keepSKL
preserves the symmetric Kullback-Leibler (SKL) divergences
- New objective function
objective.keepWADIST
preserves the weighted Aitchison distances
- Update information objectives
objective.keepEntropy
now recloses the amalgamation (as it should)objective.maxDEntropy
now recloses the amalgamation (as it should)
- New backend information functions
shannon
function calculates Shannon's indexuniformity
function calculates a normalized Shannon's index
- New objective functions
objective.keepEntropy
preserves the relative entropy of the amalgamsobjective.maxDEntropy
maximizes the between-group difference in amalgam entropy
- Update
amalgam
functionprepareArgs
now forces a re-closure of the data
- Switch
objective.maxRDA
andobjective.maxRDA2
(again)
- Switch
objective.maxRDA
andobjective.maxRDA2
- Update
objective.maxRDA
andobjective.maxRDA2
- Now sums inertia across all RD axes (as it should)
- Update
objective.maxRDA2
method- Like
objective.maxRDA
but makes "z" the community matrix
- Like
- Update
objective.maxRDA
method- Handle rare case when constrained axes explain no variance
- Add check for "z" argument
- Have
prepareArgs
coerce "z" as adata.frame
- Update
objective.maxRDA
method- Now save the intermediate data to debug
- Update
objective.maxRDA
method- Add
tryCatch
to handle the Lapack routine 'dgesdd' error - This error occurs when the SVD fails to converge
- Add
- New
as.slr
function turns components into a set of summed log-ratios - Update
amalgam
- New
asSLR
argument passed to objective functions - Objective functions now support SLRs
- Update
plot
method for SLRs
- New
- Update
amalgam
- Zeros now replaced with
zCompositions::cmultRepl
- Correctly coerce input as matrix
- Zeros now replaced with
- New scalar product methods
- New
adot
computes Aitchison scalar product for two vectors - New
amat
computes Aitchison scalar product for two matrices - New unit tests
- New
- Update
plot.amalgam
- Add argument to select which amalgams to show in ternary plot
- Move
zCompositions
to Suggests
- New
int
methodsint.fromNbit
converts a binary vector into integers
- New
weight
methodsweight.Nto1
builds a many-to-one amalgamation matrixweight.NtoN
builds a many-to-many amalgamation matrix
- New
objective
methodsobjective.keepDist
seeks to preserve Aitchison distanceobjective.maxRDA
seeks to maximize constraint of RDA
- New
amalgam
wrapper- This function runs a genetic algorithm for the objective
print.amalgam
shows the object structureplot.amalgam
makes 4-panel figure