CoupledMWCAParams-class {mwTensor}

Class "CoupledMWCAParams"

Description

The parameter object to be specified against CoupledMWCA function.

Objects from the Class

Objects can be created by calls of the form new("CoupledMWCAParams", ...).

Slots

MWCAParams has four settings as follows. For each setting, the list must have the same structure.

1. Data-wise setting Each item must be a list object that is as long as the number of data and is named after the data.

Xs:

A list containing multiple high-dimensional arrays.

mask:

A list containing multiple high-dimensional arrays, in which 0 or 1 values are filled to specify the missing elements.

pseudocount:

The pseudo count to avoid zero division, when the element is zero (Default: Machine Epsilon).

weights:

A list containing multiple high-dimensional arrays, in which some numeric values are specified to weigth each data.

2. Common Model setting Each item must be a nested list object that is as long as the number of data and is named after the data.

common_model:

Each element of the list must be a list corresponding the dimention name of data and common factor matrices name.

3. Common Factor matrix-wise setting Each item must be a list object that is as long as the number of common factor matrices and is named after the factor matrices.

common_initial:

The initial values of common factor matrices. If nothing is specified, random matrices are used.

common_algorithms:

Algorithms used to decompose the matricised tensor in each mode.

common_iteration:

The number of iterations.

common_decomp:

If FALSE is specified, unit matrix is used as the common factor matrix.

common_fix:

If TRUE is specified, the common factor matrix is not updated in the iteration.

common_dims:

The lower dimension of each common factor matrix.

common_transpose:

Whether the common factor matrix is transposed to calculate core tensor.

common_coretype:

If "CP" is specified, all the core tensors become diagonal core tensors. If "Tucker" is specified, all the core tensors become dense core tensors.

4. Specific Model setting Each item must be a nested list object that is as long as the number of data and is named after the data.

specific_model:

Each element of the list must be a list corresponding the dimention name of data and data specific factor matrices name.

5. Specific Factor matrix-wise setting Each item must be a list object that is as long as the number of data specific factor matrices and is named after the factor matrices.

specific_initial:

The initial values of data specific factor matrices. If nothing is specified, random matrices are used.

specific_algorithms:

Algorithms used to decompose the matricised tensor in each mode.

specific_iteration:

The number of iterations.

specific_decomp:

If FALSE is specified, unit matrix is used as the data specific factor matrix.

specific_fix:

If TRUE is specified, the data specific factor matrix is not updated in the iteration.

specific_dims:

The lower dimension of each data specific factor matrix.

specific_transpose:

Whether the data specific factor matrix is transposed to calculate core tensor.

specific_coretype:

If "CP" is specified, all the core tensors become diagonal core tensors. If "Tucker" is specified, all the core tensors become dense core tensors.

6. Other option Each item must to be a vector of length 1.

specific:

Whether data specific factor matrices are also calculated.

thr:

The threshold to stop the iteration. The higher the value, the faster the iteration will stop.

viz:

Whether the output is visualized.

figdir:

When viz=TRUE, whether the plot is output in the directory.

verbose:

Whether the process is monitored by verbose messages.

Methods

CoupledMWCA

Function to peform CoupledMWCA.

See Also

CoupledMWCAResult-class, CoupledMWCA