Reference#
These pages provide a reference for repliclust.
The diagram below shows our object-oriented software architecture from a big picture perspective.
A DataGenerator depends on
one or several Archetype’s to
produce synthetic data sets. Each Archetype
generates MixtureModel’s
which, in turn, sample ready-to-use synthetic data sets.
In the diagram, each box represents an object. A solid arrow
A <– B means that B is an attribute of A, while a dashed
arrow X - -> Y means that X randomly samples instances of
Y. The objects of type
CovarianceSampler,
ClusterCenterSampler,
GroupSizeSampler,
and
DistributionMix are the
modular building blocks an archetype uses to sample mixture models.
The API reference below allows you to look up the definitions of individual classes and functions in repliclust.
- Core Framework
- Max-Min Implementation
- repliclust.maxmin
- repliclust.maxmin.archetype
- repliclust.maxmin.covariance
MaxMinCovarianceSamplerMaxMinCovarianceSampler.aspect_refMaxMinCovarianceSampler.aspect_maxminMaxMinCovarianceSampler.radius_maxminMaxMinCovarianceSampler.make_axis_lengths()MaxMinCovarianceSampler.make_cluster_aspect_ratios()MaxMinCovarianceSampler.make_cluster_radii()MaxMinCovarianceSampler.sample_covariances()MaxMinCovarianceSampler.validate_k()
- repliclust.maxmin.groupsizes
- Cluster Overlap Control
- Probability Distributions