PredictEnsemble: Ensemble generation function.

In bvnlab/SCATTome: FluidgmR

Description

This function generates the enseble prediction of gene response.

Usage

PredictEnsemble(train, test, ImportantGenes, GeneNumbers = 7)

Arguments

train
test
ImportantGenes
GeneNumbers

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (train, test, ImportantGenes, GeneNumbers = 7) 
{
    dtrain <- train[, -1]
    dtest <- test[, -1]
    testcells <- test[, 1]
    library(randomForest)
    library(e1071)
    trcol <- colnames(dtrain)
    tecol <- colnames(dtest)
    SelectedSet <- ImportantGenes[1:GeneNumbers]
    trind <- c()
    tsind <- c()
    for (i in 1:length(SelectedSet)) {
        trind <- c(trind, which(trcol == SelectedSet[i]))
        tsind <- c(tsind, which(tecol == SelectedSet[i]))
    }
    if (!(length(trind) == length(tsind))) {
        print("Test and Train set do not have same important genes")
        return(0)
    }
    dtr <- dtrain[, c(trind, dim(dtrain)[2])]
    dts <- dtest[, tsind]
    rf <- randomForest(as.factor(Status) ~ ., ntree = 1000, data = dtr)
    prf <- predict(rf, newdata = dts, type = "prob")
    prf1 <- prf[, 2]
    sv <- svm(as.factor(Status) ~ ., data = dtr, kernel = "radial", 
        type = "C-classification", cross = 10, probability = TRUE)
    psv <- predict(sv, newdata = dts, probability = TRUE)
    prob <- attr(psv, "probabilities")
    prob1 <- prob[, 2]
    sv1 <- svm(as.factor(Status) ~ ., data = dtr, kernel = "sigmoid", 
        type = "C-classification", cross = 10, probability = TRUE)
    psv1 <- predict(sv, newdata = dts, probability = TRUE)
    prob <- attr(psv1, "probabilities")
    prob2 <- prob[, 2]
    ens.p = (prf1 + prob1 + prob2)/3
    Prob <- as.data.frame(cbind(ID = test[, 1], Prob = ens.p))
    write.csv(Prob, "Test_Probabilities.csv", quote = FALSE, 
        row.names = FALSE)
    pdf("Predicted_Probabilities_of_Single_Cells.pdf")
    plot(seq(1, length(ens.p)), ens.p, type = "n", ylim = c(0, 
        1), main = "Probability for Test Data")
    text(x = seq(1, length(ens.p), 1), y = ens.p, labels = test$ID, 
        col = ceiling(ens.p * 3), cex = 0.5)
    plot(seq(1, length(ens.p)), prf1, type = "n", ylim = c(0, 
        1), main = "Probability for Test Data (From Random Forest)")
    text(x = seq(1, length(ens.p), 1), y = prf1, labels = test$ID, 
        col = ceiling(ens.p * 3), cex = 0.5)
    plot(seq(1, length(ens.p)), prob1, type = "n", ylim = c(0, 
        1), main = "Probability for Test Data (From SVM 1)")
    text(x = seq(1, length(ens.p), 1), y = prob1, labels = test$ID, 
        col = ceiling(ens.p * 3), cex = 0.5)
    plot(seq(1, length(ens.p)), prob2, type = "n", ylim = c(0, 
        1), main = "Probability for Test Data (From SVM 2)")
    text(x = seq(1, length(ens.p), 1), y = prob2, labels = test$ID, 
        col = ceiling(ens.p * 3), cex = 0.5)
    dev.off()
    return(0)
  }

bvnlab/SCATTome documentation built on May 13, 2019, 9:05 a.m.