Description Usage Arguments Details Value Examples

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

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, intseed = 1, ...)
## S3 method for class 'msma'
print(x, ...)
## S3 method for class 'msma'
plot(x, ...)
``` |

`X` |
a matrix or list of matrices indicating the explanatory variable(s). This parameter is required. |

`...` |
further arguments passed to or from other methods. |

`Y` |
a matrix or list of matrices indicating objective variable(s). This is optional. If there is no input for Y, then PCA is implemented. |

`Z` |
a vector, response variable(s) for implementing the supervised version of (multiblock) PCA or PLS. This is optional. The length of Z is the number of subjects. If there is no input for Z, then 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 a length equal to the number of blocks. If lambdaX is omitted, no regularization is conducted. |

`lambdaY` |
numeric vector of regularized parameters for Y, with a length equal to the number of blocks. If lambdaY is omitted, no regularization is conducted. |

`eta` |
numeric scalar indicating the parameter indexing the penalty family. This version contains only choice 1. |

`type` |
a character, indicating the penalty family. In this version, only one choice is available: "lasso." |

`inX` |
a vector or list of numeric vectors specifying the variables in X, always included in the model |

`inY` |
a vector or list of numeric vectors specifying the variables in Y, always included in the model |

`muX` |
a numeric scalar for the weight of X for the supervised case. 0 <= muX <= 1. |

`muY` |
a numeric scalar for the weight of Y for the supervised case. 0 <= muY <= 1. |

`defmethod` |
a character representing the deflation method. This version has only the choice "canonical." |

`scaling` |
a logical, indicating whether or not data scaling is performed. The default is TRUE. |

`verbose` |
information |

`intseed` |
seed number for the random number in the parameter estimation algorithm. |

`x` |
an object of class " |

`msma`

requires at least one input X (a matrix or list). In this case, (multiblock) PCA is conducted. If Y is also specified, then a PLS is conducted using X as explanatory variables and Y as objective variables. This function scales each data matrix to a mean of 0 and variance of 1 in the default. The block structure can be represented as a list. If Z is also specified, a 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 a positive lambdaX or lambdaY is specified, then a sparse estimation based on the L1 penalty is implemented.

`dmode` |
Indicates mode "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 mean and standard deviation values for each block of X are returned. |

`Yscale` |
Scaling information for Y. The mean and standard deviation values for each block of Y are returned. |

`comp` |
Number of components |

`wbX` |
Block loading for X. The list has the same length as that of the input list X (number of blocks) and consists of a matrix. The number of variables is present in the row and the number of components is present in the column. |

`sbX` |
Block score for X. The list has the same length as that of the input list X (number of blocks) and consists of a 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 (number of blocks) and consists of a 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 (number of blocks) and consists of a matrix, with the number of subjects in the row and the number of components in the column. |

`ssX` |
Super score for X. In the matrix, the number of subjects is in the row and the number of components is in the column. |

`wsX` |
Super loading for X. In the matrix, the number of blocks is in the row and the number of components is in the column. |

`ssY` |
Super score for Y. In the matrix, the number of subjects is in the row and the number of components is in the column. |

`wsY` |
Super loading for Y. In the matrix, the number of blocks is in the row and the number of components is 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 from the original names of X |

`selectYnames` |
Names of selected variables for Y. This returns from the original names of Y |

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 March 18, 2018, 2:19 p.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.