Usage Arguments Value Author(s) Examples
1 |
DList |
Input data set matrix (a list of multiple datasets; row=features, column=samples). |
method |
"SSC" = Sum of Squared Cosine, "SV" = Sum of variance |
Meta.Dim |
Dimension size of meta-eigenvector matrix |
is.auto.Dim |
Logical value whether dimension size of each study's eigenvector matrix (SSC) is determined by an arbitrary variance quantile |
is.equal.Dim |
Logical value whether dimension size of each study's eigenvector matrix (SSC) is equal across studies |
e.Dim |
Dimension size of each study's eigenvector matrix (SSC) when is.equal.Dim = TRUE |
is.weight |
Logical value whether the reciprocal of the largest eigenvalue is mutiplied to covariance matrix |
.var.quantile |
A threshold indicating the minimum variance of individual study, when is.auto.Dim = TRUE |
.scaleAdjust |
Logical value whether the PC projection is scaled to mean of zero and SD of 1 |
is.sparse |
Logical value whether meta-eigenvector matrix is penalized to encourage sparseness |
CV_lambda |
A set of candidate tuning parameters in which the best tuning parameter is chosen |
is.plot |
Logical value whether visual scree plot is created |
The value of best tuning parameter selected among considered parameters
SungHwan Kim swiss747@gmail.com
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 | library(yeastCC)
data(yeastCC)
data<-Biobase::exprs(yeastCC)
library(impute)
library(doMC)
data.na<-is.na(data)
data.na.length<-apply(data.na, 1, sum)
data.sd<-apply(as.matrix(data), 1, sd, na.rm=TRUE)
new.data<-data[data.na.length<77*0.1 & data.sd>0.45,]
Spellman <- list(alpha=impute.knn(new.data[,5:22])$data,
cdc15=impute.knn(new.data[,23:46])$data,
cdc28=impute.knn(new.data[,47:63])$data,
elu=impute.knn(new.data[,64:77])$data)
####################################################################################################
## Searching the optimal tuning parameter based on the proportion of increased explained variance
####################################################################################################
optimal.lambda <- meta.pca.cv(DList=Spellman, method="SSC", Meta.Dim=2, CV_lambda = seq(1,10,1), is.plot=TRUE)
## optimal.lambda = 8
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.