cvxbiclustr (version 0.0.1)

Convex Biclustering Algorithm

Description

An iterative algorithm for solving a convex formulation of the biclustering problem.

Copy Link

Version

Install

install.packages('cvxbiclustr')

Monthly Downloads

Version

0.0.1

License

MIT + file LICENSE

Maintainer

Eric Chi

Last Published

June 28th, 2015

Functions in cvxbiclustr (0.0.1)

kernel_weights

Compute Gaussian Kernel Weights

create_edge_incidence

Edge-Incidence Matrix of Weights Graph

get_subgroup_means

Get Subgroup Means

compactify_edges

Construct indices matrices

knn_weights

"Thin" a weight vector to be positive only for its k-nearest neighbors

get_subgroup_means_full

Get Subgroup Means

create_adjacency

Create adjacency matrix from V

cobra_pod

MM algorithm for Convex Biclustering with Missing Data

Convert weights, as COO matrix, to CSC matrix and weights vector

gkn_weights

Gaussian Kernel + k-Nearest Neighbor Weights

cobra_validate

Perform validation to select regularization parameter

cobra

Convex Biclustering Algorithm

get_validation

Select Validation Set for a Matrix

weights_graph

Weights Graph Adjacency Matrix

cobra_internal

Convex Biclustering via Dykstra-like Proximal Algorithm

lung

Lung cancer data