Some example data for R package TransGraph
Some example data for Transfer learning of vector-valued precision matrix (undirected graphical model).
A list including:
target.DAG.data: The target DAG and data.
target.DAG.data$true_adjace: the target adjacent matrix, and we consider a DAG consisting of two equal-sized disjoint scale-free subgraphs;
target.DAG.data$Layer_true: the target topological layer structure;
target.DAG.data$n: the target sample size;
target.DAG.data$p: the data dimensional;
target.DAG.data$X: the n * p sample matrix;
target.DAG.data$Theta.t: the target precision matrix;
target.DAG.data$noise: the n * p noise matrix;
target.DAG.data$noise.type: the type of noise.
auxiliary.DAG.data: The auxiliary DAG and data. There are K=8 auxiliary domains with one parameter-informative DAG, two node-level structure-informative DAG (corresponding to two disjoint scale-free subgraphs in the target DAG, respectively), and five non-informative DAG.
Some example data for Transfer learning of tensor graphical models. See arXiv link for details.
A list including:
t.data: Tensor data in the target domain with n=50 and (p1,p2,p3)=(10,10,10).
A.data: Tensor data in informative auxiliary domains with K=5 and nk=100.
t.Omega.true.list: A list of the true precision matrices of all modes in the target domain.
Some example data for Transfer learning of non-Gaussian DAG.
A list including:
target.DAG.data: The target DAG and data.
target.DAG.data$true_adjace: the target adjacent matrix, and we consider a DAG consisting of two equal-sized disjoint scale-free subgraphs;
target.DAG.data$Layer_true: the target topological layer structure;
target.DAG.data$n: the target sample size;
target.DAG.data$p: the data dimensional;
target.DAG.data$X: the n * p sample matrix;
target.DAG.data$Theta.t: the target precision matrix;
target.DAG.data$noise: the n * p noise matrix;
target.DAG.data$noise.type: the type of noise.
auxiliary.DAG.data: The auxiliary DAG and data. There are K=8 auxiliary domains with one parameter-informative DAG, two node-level structure-informative DAG (corresponding to two disjoint scale-free subgraphs in the target DAG, respectively), and five non-informative DAG (See arXiv link for details).
Some example data of a sparse hub DAG for layer-based single non-Gaussian DAG learning. The meaning of the parameters is similar to "target.DAG.data" in "example.data.DAG.RData".