Parameter names made consistent in several auxiliary functions:
retrieveOutput()
calcQuantiles()
plotMGDrivESingle()
Minor Changes
Several spelling errors addressed.
Realistic landscape example changed to use the zero-inflated kernel,
instead of the basic exponential kernel.
Several plotting tweaks.
Added links to the accompanying data analysis package, MoNeT-MGDrivE, part
of the MoNeT
package.
MGDrivE 1.5.0
Major Changes
Complete internal rebuild of MGDrivE implementation.
The underlying mathematics are the same, only the implementation has
been changed.
Significant memory reductions.
Significant computational reduction.
All internal objects and functions have been updated.
Most important, the stochastic implementation had a bug in it, which
has been resolved.
All releases are now numeric vectors/matrices indicating the
genotype and release number; this is handled internally.
It is now possible to release mated females.
Male-mating ability is now female-genotype dependent, allowing for
assortative mating.
parameterizeMGDrivE() takes several new parameters
sampTime indicates how often output is written from the
simulation
inheritanceCube is now required to parameterize the
initial genotype distributions
LarPopRatio, AdPopRatio_F, and
AdPopRatio_M have been updated internally to reflect
accurate default behavior and handle different user input to set
them.
data.table has been removed from the dependencies. This implies
rebuilds of:
splitOutput()
aggregateFemales()
calcQuantiles()
splitOutput() and aggregateFemales() have
been fixed to properly use the writeDir parameter.
Minor Changes
All verbose options have been updated for consistency.
The default is TRUE.
All parameters that previously had to be vectors the same length as
the number of patches have been updated to take a single number,
implying that all parameters are the same for each patch, or as a
vector, so each patch can be specified individually.
Plotting functions have been updated to handle any sampling scheme
(i.e., if output is not written every day).
parameterizeMGDrivE() had internal loops replaced with
vectorized functions.
Spelling errors and documentation inconsistencies were
addressed.
Migration is handled at the Network level now. This removes objects
from the Patch class, making the package lighter and more
efficient.
Migration is no longer Dirichlet distributed.
New Inheritance Patterns
One and Two locus Cleave and Rescue (ClvR) constructs have been made
available.
ERACR/eCHACR constructs have been made available.
Minor Changes
Spelling checked and errors corrected.
Function links in the documentation have been updated.
MGDrivE 1.6.1
Major Changes
Rebuild of oneDay_mating_deterministic_Patch() and
oneDay_mating_stochastic_Patch(). They properly account for
zero probabilities with males and the associated female deaths.
This fixes a bug introduced when eta() was turned into
a matrix object, allowing selective mating.
omega() is supposed to be a matrix, and change for each
node. That isn’t implemented here, and I don’t know why not.
Minor Changes
calcOmega() has been updated internally.
A closed-form expression was derived, and the
stats::uniroot() dependency is removed.
This does not actually fix the issue found, which is in the bounds
for mu.
Better bounds-checking was introduced, with a more clear error
message, to warn users when they encounter this error.
MGDrivE 1.6.2
Major Changes
None
Minor Changes
Fixed deposition issues in cubeClvR2() - Thank you
Dr. Leftwich.
Updated internal documentation - per CRAN requests.
Exported several new inheritance patterns that have been in
development.
Updated website documentation, switched to the current version of
pkgdown.