my_PCA — ds-pca • omXplore

Process a PCA, using nipals or FactoMineR, on a quantitative dataset.

This method plots a bar plot which represents the distribution of the number of missing values (NA) per lines (ie proteins).

wrapper_pca()
plotPCA_Eigen_hc(): plots the eigen values of PCA with the highcharts library
plotPCA_Eigen(): plots the eigen values of PCA
plotPCA_Var()
plotPCA_Ind()

my_PCA(
  X,
  scale.unit = TRUE,
  ncp = min(12, nrow(X) - 1, ncol(X)),
  ind.sup = NULL,
  quanti.sup = NULL,
  quali.sup = NULL,
  row.w = NULL,
  col.w = NULL,
  graph = FALSE,
  axes = c(1, 2),
  method = "FM",
  gramschmidt = TRUE
)

omXplore_pca_ui(id)

omXplore_pca_server(id, obj, i)

omXplore_pca(obj, i)

wrapper_pca(
  qdata,
  group,
  var.scaling = TRUE,
  ncp = NULL,
  method = NULL,
  gramschmidt = TRUE
)

plotPCA_Eigen(res.pca)

plotPCA_Var(res.pca, chosen.axes = c(1, 2))

plotPCA_Ind(res.pca, chosen.axes = c(1, 2))

plotPCA_Eigen_hc(res.pca)

Arguments

X: a data.frame() of quantitative data
scale.unit: See FactoMineR::PCA()
ncp: See FactoMineR::PCA()
ind.sup: See FactoMineR::PCA()
quanti.sup: See FactoMineR::PCA()
quali.sup: See FactoMineR::PCA()
row.w: See FactoMineR::PCA()
col.w: See FactoMineR::PCA()
graph: See FactoMineR::PCA()
axes: See FactoMineR::PCA()
method: a string corresponding to the package to use for PCA (if no NA, default is "FM" for FactoMineR)
gramschmidt: A boolean indicating whether to use Gram-Schmidt orthogonalization or not.
id: A character(1) which is the id of the shiny module.
obj: An instance of the class MultiAssayExperiment.
i: An integer which is the index of the assay in the param obj
qdata: A data.frame() of quantitative data
group: A vector with the name of samples
var.scaling: A boolean indicating whether to scale the data or not
res.pca: The result of the function FactoMineR::PCA()
chosen.axes: See the parameter 'axes' of the function factoextra::fviz_pca_var()

Value

res.pca a "PCA" "list" object

A shiny app

The result of the function FactoMineR::PCA()

A plot

Author

Samuel Wieczorek, Enora Fremy

Examples

data(vdata)
obj <- vdata[[1]]
res.pca <- my_PCA(SummarizedExperiment::assay(obj), method = "FM")
#> [1] "FM"
plotPCA_Eigen(res.pca)

plotPCA_Var(res.pca)

plotPCA_Eigen_hc(res.pca)
plotPCA_Ind(res.pca)


if (FALSE) { # \dontrun{
  data(vdata)
  library(shinyWidgets)
  library(dplyr)
  # Replace missing values for the example
  sel <- is.na(SummarizedExperiment::assay(vdata, 1))
  SummarizedExperiment::assay(vdata[[1]])[sel] <- 0
  omXplore_pca(vdata, 1)
  shiny::runApp(omXplore_pca(vdata, 1))
} # }


data(vdata)
obj <- vdata[[1]]
res.pca <- wrapper_pca(SummarizedExperiment::assay(obj), get_group(obj))
#> NULL
#> Error in if (method == "NIPALS") X.init <- X: argument is of length zero
plotPCA_Eigen(res.pca)

plotPCA_Var(res.pca)

plotPCA_Eigen_hc(res.pca)
plotPCA_Ind(res.pca)