`MapperRef` is an R6 class built for parameterizing and computing mappers efficiently.
MapperRef instance equipped with methods for building the mapper.
To create a new MapperRef instance object, instantiate the class with the R6 new operator.
Instantiation of a MapperRef objects requires a data matrix, or a function returning one.
The primary output of the Mapper method is a simplicial complex.
The underlying complex does not need to be modified by the user, i.e. is completely maintained by MapperRef methods
(e.g. construct_nerve, construct_k_skeleton, etc.).
m = MapperRef$new(X)
X: The data, either as a matrix, or a function that returns a matrix.
X: The data, as a function. Evaluation returns the data as a matrix.
filter: The filter, as a function. Evaluation returns the filtered data as a matrix.
cover: The cover (a CoverRef derived object).
clustering_algorithm: The clustering algorithm to use in the pullback.
measure: Distance measure to use to compute distances in ambient space. See use_distance_measure for more details.
pullback: Mapping between the sets in the cover (by index) and the vertices (by id).
vertices: The mapper vertices.
simplicial_complex: A simplex tree object.
use_filter: Specifies the filter.
use_distance_measure: Specifies the distance measure.
use_clustering_algorithm: Specifies the algorithm to decompose the pullback.
construct_pullback: Decomposes the preimages into connected components.
construct_nerve: Constructs simplices at a given dimension.
construct_k_skeleton: Constructs the simplicial complex up to a given dimension.
as_igraph: Converts the 1-skeleton to an igraph object.
as_pixiplex: Converts the simplicial complex to a pixiplex object.
exportMapper: Exports the core information of the mapper construction.
Full documentation available online.
Gurjeet Singh, Facundo Mémoli, and Gunnar Carlsson. "Topological methods for the analysis of high dimensional data sets and 3d object recognition." SPBG. 2007.