stochprofML 2.0.3
- export function set.model.function() as it is needed when using for
example d.sum.of.mixtures.
- in stochprof.results generate a duplicate of all previous results
perform rounding of parameters and target and remove duplicates in the
original result table, like this all results in the final output and
used inside optimization are not rounded and belong 100% to the target
negative loglikelihodd and BIC.
stochprofML 2.0.2
- d.sum.of.lognormals: If it is not a real sum but only one summand
use dlnorm directly
- in d.sum.of.types (all models) and correspondingly
d.sum.of.lognormal.types: bug fix, as mu.vector and sigma.vector were
wrongly filled for TY > 2.
- small foramting changes on Help pages
stochprofML 2.0.1
- Deleted the argument “logdens” in mix.d.sum.of.mixtures because of a
bug if set to TRUE.
stochprofML 2.0.0
- Added a
NEWS.md
file to track changes to the
package.
- Created a github repository:
https://github.com/fuchslab/stochprofML
- Extended n the number of cells that was constant over all samples to
be flexible to be different for each sample.
- Added possibility to put in the predefined number of cells when
using the random number generators r.sum.of.mixtures by specifying
N.matrix
- Add the mixed density calculator when calculating the density of a
distribution with a specific n.vector as input.
- Changed all functions that use the fixed n to be able to use a
vector n
- Added doi of Tirier et al. (2019) paper where this was first
implemented and used
- Minor fixes of textings