# msma: Multiblock Sparse Multivariable Analysis In msma: Multiblock Sparse Multivariable Analysis

## Description

This is a function for a matrix decomposition method incorporating sparse and supervised modeling for a multiblock multivariable data analysis

## Usage

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13``` ```msma(X, ...) ## Default S3 method: msma(X, Y = NULL, Z = NULL, comp = 2, lambdaX = NULL, lambdaY = NULL, eta = 1, type = "lasso", inX = NULL, inY = NULL, muX = 0, muY = 0, defmethod = "canonical", scaling = TRUE, verbose = FALSE, ...) ## S3 method for class 'msma' print(x, ...) ## S3 method for class 'msma' plot(x, ...) ```

## Arguments

 `X` a (list of) matrix, explanatory variable(s) which is required. `...` further arguments passed to or from other methods. `Y` a (list of) matrix, objective variable(s). This is optional. If no input for Y, then the PCA method is implemented. `Z` a vector, response variable(s). This is optional. The length is the number of subjects. If no input for Z, then the unsupervised PLS/PCA is implemented. `comp` numeric scalar for the number of components to be considered. `lambdaX` numeric vector of regularized parameters for X with length equal to the number of blocks. If omitted, no regularization is conducted. `lambdaY` numeric vector of regularized parameters for Y with length equal to the number of blocks. If omitted, no regularization is conducted. `eta` numeric scalar, the parameter indexing the penalty family. This version has only the choice 1. `type` a character, the penalty family. This version has only the choice "lasso". `inX` a (list of) numeric vector to specify the variables of X which are always in the model. `inY` a (list of) numeric vector to specify the variables of Y which are always in the model. `muX` a numeric scalar for the weight of X for the supervised. 0<=muX<=1. `muY` a numeric scalar for the weight of Y for the supervised. 0<=muY<=1. `defmethod` a character, the deflation method, this version has only the choice "canonical". `scaling` a logical, whether the scaling data is done, the default is TRUE. `verbose` information `x` an object of class "`msma`", usually, a result of a call to `msma`

## Details

`msma` requires at least one input X as the matrix or the list. In this case, the (multiblock) principal components analysis is conducted. If Y is also specified, the partial least squares with X as explanatory variables and Y as objective variables. This function scaled each data matrix to mean 0 and variance 1 in the default. The block structure can be represented as the list. If Z is also specified, the supervised version is implemented and the degree is controlled by muX or muY where 0<=muX<=1, 0<=muY<=1, and 0<=muX+muY<1. If the positive lambdaX or lambdaY is specified, the sparse estimation based on L1 penalty is implemented.

## Value

 `dmode` Which modes "PLS" or "PCA" `X` Scaled X which has a list form. `Y` Scaled Y which has a list form. `Xscale` Scaling information for X. The means and standard deviations for each block of X are returned. `Yscale` Scaling information for Y. The means and standard deviations for each block of Y are returned. `comp` the number of components `wbX` block loading for X. The list has same length as that of the input list X (the number of blocks) and consists of the matrix with the number of variables in the row and the number of components in the column. `sbX` block score for X. The list has same length as that of the input list X (the number of blocks) and consists of the matrix with the number of subjects in the row and the number of components in the column. `wbY` block loading for Y. The list has same length as that of the input list Y (the number of blocks) and consists of the matrix with the number of variables in the row and the number of components in the column. `sbY` block score for Y. The list has same length as that of the input list Y (the number of blocks) and consists of the matrix with the number of subjects in the row and the number of components in the column. `ssX` super score for X. The matrix has the number of subjects in the row and the number of components in the column. `wsX` super loading for X. The matrix has the number of blocks in the row and the number of components in the column. `ssY` super score for Y. The matrix has the number of subjects in the row and the number of components in the column. `wsY` super loading for Y. The matrix has the number of blocks in the row and the number of components in the column. `nzwbX` number of nonzeros in block loading for X `nzwbY` number of nonzeros in block loading for Y `selectXnames` names of selected variables for X. This returns the names of X `selectYnames` names of selected variables for Y. This returns the names of Y

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24``` ```##### data ##### tmpdata = simdata(n = 50, rho = 0.8, Yps = c(10, 12, 15), Xps = 20, seed=1) X = tmpdata\$X; Y = tmpdata\$Y ##### One Component ##### fit1 = msma(X, Y, comp=1, lambdaX=2, lambdaY=1:3) fit1 ##### Two Component ##### fit2 = msma(X, Y, comp=2, lambdaX=2, lambdaY=1:3) fit2 ##### Matrix data ##### sigma = matrix(0.8, 10, 10) diag(sigma) = 1 X2 = rmvnorm(50, rep(0, 10), sigma) Y2 = rmvnorm(50, rep(0, 10), sigma) fit3 = msma(X2, Y2, comp=1, lambdaX=2, lambdaY=2) fit3 ##### Sparse Principal Component Analysis ##### fit5 = msma(X2, comp=5, lambdaX=2.5) summary(fit5) ```

msma documentation built on May 30, 2017, 2:27 a.m.