some questions about the function reliability.test()

[ZiXuan0606]

I used the colorectal cancer sc_dataset and bulk data.I uesd the DFI infomation and cox regression.But when I used the function reliability.test(),the pvalue is na,and sometimes statistic is nan,I check the function,and find that due to sample distribution,I sample the same status all is 0.So I want to ask you how to solve this problem.I choose family = binomial or modify the function test_cox according to the annotation,annotating code index0,choose index1 and index2 as following.Scissor is a very important tool for me,thank you very much .I also sent email to you,sorry to bother you. test_cox <- function(X, Y, network, alpha, cell_num, n = 100, nfold = 10){ set.seed(1) m1 <- sum(Y[,2] == 1) m2 <- sum(Y[,2] == 0)

index0 <- sample(cut(seq(m1+m2), breaks = 10, labels = F))

index1 <- sample(cut(seq(m1), breaks = nfold, labels = F))
index2 <- sample(cut(seq(m2), breaks = nfold, labels = F))

print("|**************************************************|")
print("Perform cross-validation on X with true label")
c_index_test_real <- NULL
pb1 <- progress_bar$new(total = nfold)
for (j in 1:nfold){
    #c_index <- which(index0 == j)
    c_index <- c(which(Y[,2] == 1)[which(index1 == j)], which(Y[,2] == 0)[which(index2 == j)])######I changed here
    X_train <- X[-c_index,]
    Y_train <- Y[-c_index,]
    fit <- NULL
    while (is.null(fit$fit)){
        set.seed(123)
        fit <- APML1(X_train, Y_train, family = "cox", penalty = "Net", alpha = alpha, Omega = network, nlambda = 100)
    }
    index <- which.min(abs(fit$fit$nzero - cell_num))
    Coefs <- as.numeric(fit$Beta[,index])
    Cell1 <- Coefs[which(Coefs > 0)]
    Cell2 <- Coefs[which(Coefs < 0)]

    X_test <- X[c_index,]
    Y_test <- Y[c_index,]
    test_data <- data.frame(cbind(Y_test, X_test%*%Coefs))
    colnames(test_data) <- c("OS_time", "Status", "Prediction")
    res.cox <- coxph(Surv(OS_time, Status) ~ Prediction, data = test_data)
    c_index_test_real[j] <- concordance(res.cox)$concordance

    pb1$tick()
    Sys.sleep(1 / 100)
    if (j == nfold) cat("Finished!\n")
}

print("|**************************************************|")
print("Perform cross-validation on X with permutated label")
c_index_test_back <- list()
pb2 <- progress_bar$new(total = n)
for (i in 1:n){
    set.seed(i+100)
    c_index_test_back[[i]] <- matrix(0, nfold, 1, dimnames = list(paste0("Testing_", 1:nfold),  "Concordance"))
    Y2 <- Y[sample(nrow(Y)),]
    for (j in 1:nfold){
        #c_index <- which(index0 == j)
        c_index <- c(which(Y2[,2] == 1)[which(index1 == j)], which(Y2[,2] == 0)[which(index2 == j)])######I changed here
        X_train <- X[-c_index,]
        Y_train <- Y2[-c_index,]
        fit <- NULL
        while (is.null(fit$fit)){
            set.seed(123)
            fit <- APML1(X_train, Y_train, family = "cox", penalty = "Net", alpha = alpha, Omega = network, nlambda = 100)
        }
        index <- which.min(abs(fit$fit$nzero - cell_num))
        Coefs <- as.numeric(fit$Beta[,index])
        Cell1 <- Coefs[which(Coefs > 0)]
        Cell2 <- Coefs[which(Coefs < 0)]

        X_test <- X[c_index,]
        Y_test <- Y2[c_index,]
        test_data <- data.frame(cbind(Y_test, X_test%*%Coefs))
        colnames(test_data) <- c("OS_time", "Status", "Prediction")
        res.cox <- coxph(Surv(OS_time, Status) ~ Prediction, data = test_data)
        c_index_test_back[[i]][j] <- concordance(res.cox)$concordance
    }
    pb2$tick()
    Sys.sleep(1 / 100)
    if (i == n) cat("Finished!\n")
}
statistic  <- mean(c_index_test_real)
background <- NULL
for (i in 1:n){
    background[i] <- mean(c_index_test_back[[i]][,1])
}
p <- sum(background > statistic)/n

print(sprintf("Test statistic = %s", formatC(statistic, format = "f", digits = 3)))
print(sprintf("Reliability significance test p = %s", formatC(p, format = "f", digits = 3)))

return(list(statistic = statistic,
            p = p,
            c_index_test_real = c_index_test_real,
            c_index_test_back = c_index_test_back))

}

[sunduanchen]

Problem solved.

[oyxf]

I have the same questions , and my paramter's setting as follow： Scissor::reliability.test(X, Y, network, alpha =alpha_param[i], family = "gaussian", cell_num = numbers, n = 10, nfold = 100)。 Could you tell me how to solve this problem? Thank you very much..

[elfthrith]

I have the same questions , and my paramter's setting as follow： Scissor::reliability.test(X, Y, network, alpha =alpha_param[i], family = "gaussian", cell_num = numbers, n = 10, nfold = 100)。 Could you tell me how to solve this problem? Thank you very much..

Hi, I also have this problem. The parameter setting is the same as yours. Have you solved this problem? If you could reply, i would appreciate.