plot,oplsMultiDataSet,ANY-method {ropls} | R Documentation |
This function plots values based upon a model trained by opls
.
This function plots values based upon a model trained by opls
.
## S4 method for signature 'oplsMultiDataSet,ANY' plot( x, y, fig.pdfC = c("none", "interactive", "myfile.pdf")[2], info.txtC = c("none", "interactive", "myfile.txt")[2], ... ) ## S4 method for signature 'opls,ANY' plot( x, y, typeVc = c("correlation", "outlier", "overview", "permutation", "predict-train", "predict-test", "summary", "x-loading", "x-score", "x-variance", "xy-score", "xy-weight")[7], parAsColFcVn = NA, parCexN = 0.8, parCompVi = c(1, 2), parEllipsesL = NA, parLabVc = NA, parPaletteVc = NA, parTitleL = TRUE, parCexMetricN = NA, plotPhenoDataC = NA, plotSubC = NA, fig.pdfC = c("none", "interactive", "myfile.pdf")[2], info.txtC = c("none", "interactive", "myfile.txt")[2], file.pdfC = NULL, .sinkC = NULL, ... )
x |
An S4 object of class |
y |
Currently not used |
fig.pdfC |
Character: File name with '.pdf' extension for the figure; if 'interactive' (default), figures will be displayed interactively; if 'none', no figure will be generated |
info.txtC |
Character: File name with '.txt' extension for the printed results (call to sink()'); if 'interactive' (default), messages will be printed on the screen; if 'none', no verbose will be generated |
... |
Currently not used. |
typeVc |
Character vector: the following plots are available: 'correlation': Variable correlations with the components, 'outlier': Observation diagnostics (score and orthogonal distances), 'overview': Model overview showing R2Ycum and Q2cum (or 'Variance explained' for PCA), 'permutation': Scatterplot of R2Y and Q2Y actual and simulated models after random permutation of response values; 'predict-train' and 'predict-test': Predicted vs Actual Y for reference and test sets (only if Y has a single column), 'summary' [default]: 4-plot summary showing permutation, overview, outlier, and x-score together, 'x-variance': Spread of raw variables corresp. with min, median, and max variances, 'x-loading': X-loadings (the 6 of variables most contributing to loadings are colored in red to facilitate interpretation), 'x-score': X-Scores, 'xy-score': XY-Scores, 'xy-weight': XY-Weights |
parAsColFcVn |
Optional factor character or numeric vector to be converted into colors for the score plot; default is NA [ie colors will be converted from 'y' in case of (O)PLS(-DA) or will be 'black' for PCA] |
parCexN |
Numeric: amount by which plotting text should be magnified relative to the default |
parCompVi |
Integer vector of length 2: indices of the two components to be displayed on the score plot (first two components by default) |
parEllipsesL |
Should the Mahalanobis ellipses be drawn? If 'NA' [default], ellipses are drawn when either a character parAsColVcn is provided (PCA case), or when 'y' is a character factor ((O)PLS-DA cases). |
parLabVc |
Optional character vector for the labels of observations on the plot; default is NA [ie row names of 'x', if available, or indices of 'x', otherwise, will be used] |
parPaletteVc |
Optional character vector of colors to be used in the plots |
parTitleL |
Should the titles of the plots be printed on the graphics (default = TRUE); It may be convenient to set this argument to FALSE when the user wishes to add specific titles a posteriori |
parCexMetricN |
Numeric: magnification of the metrics at the bottom of score plot (default -NA- is 1 in 1x1 and 0.7 in 2x2 display) |
plotPhenoDataC |
Character: if x was generated from an ExpressionSet (i.e. if the 'eset' slot from x is not NULL), the name of the pData(x) column to be used for coloring can be specified here (instead of 'parAsColFcVn') |
plotSubC |
Character: Graphic subtitle |
file.pdfC |
Character: deprecated; use the 'fig.pdfC' argument instead |
.sinkC |
Character: deprecated; use the 'info.txtC' argument instead |
# Loading the 'NCI60_4arrays' from the 'omicade4' package data("NCI60_4arrays", package = "omicade4") # Selecting two of the four datasets setNamesVc <- c("agilent", "hgu95") # Creating the MultiDataSet instance nciMset <- MultiDataSet::createMultiDataSet() # Adding the two datasets as ExpressionSet instances for (setC in setNamesVc) { # Getting the data exprMN <- as.matrix(NCI60_4arrays[[setC]]) pdataDF <- data.frame(row.names = colnames(exprMN), cancer = substr(colnames(exprMN), 1, 2), stringsAsFactors = FALSE) fdataDF <- data.frame(row.names = rownames(exprMN), name = rownames(exprMN), stringsAsFactors = FALSE) # Building the ExpressionSet eset <- Biobase::ExpressionSet(assayData = exprMN, phenoData = new("AnnotatedDataFrame", data = pdataDF), featureData = new("AnnotatedDataFrame", data = fdataDF), experimentData = new("MIAME", title = setC)) # Adding to the MultiDataSet nciMset <- MultiDataSet::add_eset(nciMset, eset, dataset.type = setC, GRanges = NA, warnings = FALSE) } # Summary of the MultiDataSet nciMset # Principal Component Analysis of each data set nciPca <- ropls::opls(nciMset) # Coloring the Score plot according to cancer types ropls::plot(nciPca, y = "cancer", typeVc = "x-score") # Restricting to the 'ME' and 'LE' cancer types sampleNamesVc <- Biobase::sampleNames(nciMset[["agilent"]]) cancerTypeVc <- Biobase::pData(nciMset[["agilent"]])[, "cancer"] nciMset <- nciMset[sampleNamesVc[cancerTypeVc %in% c("ME", "LE")], ] # Building PLS-DA models for the cancer type nciPlsda <- ropls::opls(nciMset, "cancer", predI = 2) data(sacurine) attach(sacurine) for(typeC in c("correlation", "outlier", "overview", "permutation", "predict-train","predict-test", "summary", "x-loading", "x-score", "x-variance", "xy-score", "xy-weight")) { print(typeC) if(grepl("predict", typeC)) subset <- "odd" else subset <- NULL plsModel <- opls(dataMatrix, sampleMetadata[, "gender"], predI = ifelse(typeC != "xy-weight", 1, 2), orthoI = ifelse(typeC != "xy-weight", 1, 0), permI = ifelse(typeC == "permutation", 10, 0), subset = subset, info.txtC = "none", fig.pdfC = "none") plot(plsModel, typeVc = typeC) } sacPlsda <- opls(dataMatrix, sampleMetadata[, "gender"]) plot(sacPlsda, parPaletteVc = c("green4", "magenta")) #### Application to an ExpressionSet sacSet <- Biobase::ExpressionSet(assayData = t(dataMatrix), phenoData = new("AnnotatedDataFrame", data = sampleMetadata), featureData = new("AnnotatedDataFrame", data = variableMetadata), experimentData = new("MIAME", title = "sacurine")) sacPlsda <- opls(sacSet, "gender") plot(sacPlsda, "gender", typeVc = "x-score") detach(sacurine)