Description Usage Arguments Value Examples
This function finds the selected groups, subgroups and individual predictors from a sgspls method.
1 | select.sgspls(model, module, gene, time)
|
model |
object of class inhereting from "sgspls". |
module |
A p-vector indicating group membership for each covariate in the X-block |
gene |
A p-vector indicating gene membership for each covariate in the X-block |
time |
A p-vector indicating time membership for each covariate in the X-block |
Returns a list with the following selected parameter information:
select.table.X |
A table detailing the number of times each gene has been selected at each timepoint and the number of consistently selected genes. |
summary.table |
A table sumarising the number of modules, genes, timepoints and covariates selected. |
tab.gene.X |
Lists the number of timepoint that each gene in a given module and component occurs. |
tab.gene.time.X |
Table of the selected genes against time points that they occur. |
consistent.genes.X |
Returns the genes that occur in every time point. |
select.gene.X |
Returns the genes that are selected at least once for a given component. |
select.gene.X.total |
Returns the genes that are selected at least once across any component. |
selected.table.gene.X |
Returns the total number of genes selected at each component. |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 | set.seed(1)
n = 50; p = 510;
size.groups = 30; size.subgroups = 5
groupX <- ceiling(1:p / size.groups)
subgroupX <- ceiling(1:p / size.subgroups)
X = matrix(rnorm(n * p), ncol = p, nrow = n)
beta <- rep(0,p)
bSG <- -2:2; b0 <- rep(0,length(bSG))
betaG <- c(bSG, b0, bSG, b0, bSG, b0)
beta[1:size.groups] <- betaG
y = X %*% beta + 0.1*rnorm(n)
# Beta contains 1 active module with 3 active genes
# index for modules, genes and times
mod_index <- groupX
gene_index <- subgroupX
time_index <- rep(rep(1:5,6), 17)
model <- sgspls(X, y, ncomp = 3, mode = "regression", keepX = 1,
groupX = groupX, subgroupX = subgroupX,
indiv_sparsity_x = 0.8, subgroup_sparsity_x = 0.15)
reg_coef <- coef(model, type = "coefficients")
# Check model fit
cbind(reg_coef$B[ , , 3], beta)
# See the estimated regression coefficient
cbind(reg_coef$B[,,3], beta, mod_index, gene_index, time_index)[1:30,]
selectedVar <- select.sgspls(model, module = mod_index, gene = gene_index, time = time_index)
# show number of selected genes for component 1
selectedVar$select.table.X$comp1
# show number of modules, genes, times and total variables selected
selectedVar$summary.table
# Show when genes were selected from given module
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.